research: add paired original vs proposed PROTEAN analysis

2026-10-09 13:40:20 +07:00
1 parent 87f9b14ada
commit ad8b2d7671
@@ -583,7 +583,79 @@ difference; the matrices come from the same manifest, split, seed, and class ord
 ![Paired confusion matrices](artifacts/figures/final/figH_confusion_matrix_5gnidd_a0.75_seed0.png)


-## 17. Main Scientific Findings
+## 17. Deep Comparison: Original PROTEAN vs Proposed (Result)
+
+A dedicated paired analysis compares `original_protean` against
+`support_weighted` at the validation-selected beta for every dataset-alpha condition, using the
+tests split and hard-coded pairing checks (same dataset, alpha, seed, and `manifest_content_hash`).
+Full report: **`docs/research/ORIGINAL_VS_PROPOSED_DEEP_COMPARISON.md`**. Derived tables:
+`artifacts/analysis/deepcmp__*.csv` (produced by `scripts/deep_comparison.py`). Figures:
+`scripts/make_comparison_figures.py`.
+
+**Global test overview** (Δ = proposed − original, mean over seeds `{0,1,2}`; rare-class macro-F1 shown
+in full, other metrics as deltas):
+
+| Dataset | alpha | beta* | rare-F1 original | rare-F1 proposed | Δ rare-F1 | Δ rare-recall | Δ macro-F1 | Δ accuracy |
+|---|---|---|---|---|---|---|---|---|
+| X-IIoTID | 0.25 | 0.00 | 0.6328 | 0.6321 | −0.0007 | −0.0007 | −0.0007 | −0.0014 |
+| X-IIoTID | 0.50 | 0.75 | 0.6425 | 0.6655 | **+0.0230** | −0.0144 | +0.0222 | +0.0069 |
+| X-IIoTID | 0.75 | 0.50 | 0.6039 | 0.6052 | +0.0014 | −0.0077 | +0.0058 | +0.0075 |
+| 5G-NIDD | 0.25 | 0.00 | 0.8423 | 0.8412 | −0.0010 | +0.0007 | −0.0028 | −0.0007 |
+| 5G-NIDD | 0.50 | 1.00 | 0.8805 | 0.8720 | −0.0086 | −0.0058 | −0.0038 | −0.0007 |
+| 5G-NIDD | 0.75 | 0.50 | 0.8799 | 0.8967 | **+0.0168** | −0.0015 | +0.0117 | +0.0013 |
+
+Both `alpha=0.25` conditions selected `beta=0` and are numerically null. Among the four genuine
+`beta>0` conditions, rare-class macro-F1 improved in three and degraded in one; overall macro-F1 moves
+with it, while rare-class recall usually falls (precision–recall trade-off). 5G-NIDD a0.50 is
+seed-dominated. No statistical significance is claimed.
+
+### Fig I — Paired per-client test rare-class macro-F1 (proposed vs ORIGINAL PROTEAN)
+
+Each point is one (client, seed) test observation, faceted by dataset and alpha; charcoal/teal/red/gray
+and markers encode outcome and seed. At `beta>0` conditions the response is client-specific (e.g.
+X-IIoTID a0.50: 12/30 improved, 1 degraded, 17 unchanged), with individual changes larger than the
+condition mean; at `beta=0` conditions almost all observations are unchanged.
+
+![Paired per-client comparison](artifacts/figures/final/figI_per_client_original_vs_proposed.png)
+
+### Fig J — Per-class test F1 change (beta>0 conditions)
+
+Signed Δ F1 per class, panels per dataset-alpha; class support is annotated. Largest gains are
+X-IIoTID `Tampering` (+0.14) and `Reconnaissance` (+0.05) and 5G-NIDD `TCPConnectScan` (+0.03) /
+`ICMPFlood` (+0.10, tiny support); largest losses are X-IIoTID `C&C` (−0.06) and 5G-NIDD `SYNFlood`
+(−0.04). Benign detection is essentially unchanged.
+
+![Per-class comparison](artifacts/figures/final/figJ_per_class_original_vs_proposed.png)
+
+### Fig K — Confusion-matrix difference (proposed − original)
+
+For the paired condition 5G-NIDD, `alpha=0.75`, `seed=0` (identical manifest, split, class order,
+243,177 test samples): left = raw count difference, right = row-normalized rate difference. The
+dominant error `Benign → UDPFlood` is essentially unchanged (56,899 → 57,054, +155). The main reduced
+errors are `HTTPFlood → SlowrateDoS` (−547) and `Benign → TCPConnectScan` (−520); the main increased
+error is `SlowrateDoS → HTTPFlood` (+229). Red = positive delta, blue = negative; no significance is
+inferred.
+
+![Confusion-matrix difference](artifacts/figures/final/figK_confusion_delta_5gnidd_a0.75_seed0.png)
+
+**Per-client findings.** The proposed method redistributes benefit: within a condition some clients
+improve and others degrade, and the low-rare-support quartile is not systematically disadvantaged
+(descriptive; Pearson Δ-vs-support-fraction ≈ −0.09, n=120). Client ids are per-seed realizations.
+
+**Per-class findings.** Rare-class gains are partly precision-driven (recall often falls) and are not
+confined to rare classes; the largest regressions are on low-support attack classes (X-IIoTID `C&C`,
+5G-NIDD `SYNFlood`).
+
+**Confusion-delta findings.** The weighting does not fix the dominant Benign→UDPFlood error; the
+measurable benefit at the analysed pair is a reduced HTTPFlood↔SlowrateDoS confusion, with a small
+SlowrateDoS recall cost.
+
+**Practical implications and limitations.** Effects are small (≤ ~2.3 pp rare-F1) and
+condition-dependent; the largest gain is comparable to seed noise. Practical benefit is modest, not a
+blanket improvement, and comes with rare-recall and some client- or class-level costs. Pairing,
+metric definitions, and threats to validity are detailed in the report.
+
+## 18. Main Scientific Findings

 1. **Changing `beta` changes the global prototype.** Contribution weights become proportional to
   support, and for a fixed snapshot mean prototype displacement grows monotonically with `beta` in
@@ -602,7 +674,7 @@ difference; the matrices come from the same manifest, split, seed, and class ord
   not a proxy for detection quality.
 8. **The dominant classification error (Benign → UDPFlood) is not addressed by `beta`.**

-## 18. Limitations and Novelty Boundary
+## 19. Limitations and Novelty Boundary

 **Limitations.**

@@ -623,7 +695,7 @@ contribution is the controlled `beta` parameterization inside PROTEAN, the valid
 condition-wise selection, the paired multi-seed design, and the behavioral findings. No claim is
 made that the method universally improves PROTEAN.

-## 19. Repository Structure
+## 20. Repository Structure

 ```
 src/protean_research/   data loading/preprocessing/partitioning, models, aggregation, federated
@@ -643,7 +715,7 @@ artifacts/figures/final/ thesis-ready figures
 results/                local run outputs (git-ignored; 108 non-smoke run dirs + 1 smoke)
 ```

-## 20. Reproduction and Tests
+## 21. Reproduction and Tests

 ```bash
 # unit + integration tests (expected: 75 passed)
@@ -682,7 +754,7 @@ rerun the grid.
  the shared `S_j`.
 - `beta` selection uses validation only; the test set is never used for selection.

-## 21. Complete Artifact Index
+## 22. Complete Artifact Index

 Per-run outputs (`results/<dataset>__alpha<alpha>__seed<seed>__<config>/`):

@@ -720,7 +792,7 @@ Analysis outputs (`artifacts/analysis/`), by function:
 | Prototype snapshots & weights | `*__prototype_snapshot.json`, `*__prototype_weights.csv`, `*__prototype_effect*.csv` |

 Partition manifests: `artifacts/partitions/<dataset>__alpha<alpha>__seed<seed>.json` (18 files).
-Figures: `artifacts/figures/final/` — 9 figures (figA–figH) as PNG (220 dpi) plus matching vector PDF;
+Figures: `artifacts/figures/final/` — 12 figures (figA–figK) as PNG (220 dpi) plus matching vector PDF;
 regenerated reproducibly by `scripts/make_final_figures.py`.

 Key documents:
@@ -729,6 +801,7 @@ Key documents:
 |---|---|
 | `docs/research/EXPERIMENT_PROTOCOL.md` | frozen reproduction protocol |
 | `docs/research/FINAL_EXPERIMENTAL_SYNTHESIS.md` | authoritative six-condition synthesis |
+| `docs/research/ORIGINAL_VS_PROPOSED_DEEP_COMPARISON.md` | paired original-vs-proposed deep comparison |
 | `docs/research/CROSS_DATASET_BETA_ANALYSIS.md` | cross-dataset comparison |
 | `docs/research/XIIOTID_CROSS_ALPHA_ANALYSIS.md` | X-IIoTID cross-alpha analysis |
 | `docs/research/5GNIDD_CROSS_ALPHA_ANALYSIS.md` | 5G-NIDD cross-alpha analysis |
@@ -741,7 +814,7 @@ Key documents:
 | `docs/handoff/SESSION_HANDOFF.md` | current state and next steps |
 | `planningawal/PROTEAN_Revisi_Lurus.md` | immutable research specification |

-## 22. References
+## 23. References

 - Chennoufi, S., Han, Y., Blanc, G., De Cristofaro, E., & Kiennert, C. (2025). PROTEAN: Federated
  Intrusion Detection in Non-IID Environments through Prototype-Based Knowledge Sharing. ESORICS
@@ -0,0 +1,121 @@
+dataset,alpha,seed,client,delta,rare_support,total,rare_frac,present_classes,missing
+xiiotid,0.5,0,0,-0.0005042259896347323,144,94564,0.001522778224271393,7,2
+xiiotid,0.5,0,1,0.008524445085661969,249,50631,0.004917935652070866,9,0
+xiiotid,0.5,0,2,0.018813433543958924,97,52476,0.0018484640597606526,9,0
+xiiotid,0.5,0,3,-0.2565947651121153,770,69609,0.011061787987185566,9,0
+xiiotid,0.5,0,4,0.014800433413590774,19,18521,0.0010258625344203877,9,0
+xiiotid,0.5,0,5,0.13628930367827585,6,41886,0.00014324595330181923,9,0
+xiiotid,0.5,0,6,0.08483385422643441,137,45085,0.003038704668958634,7,2
+xiiotid,0.5,0,7,-0.005389002939121236,30,84396,0.00035546708374804495,9,0
+xiiotid,0.5,0,8,-0.008577128576549353,49,85073,0.000575975926557192,9,0
+xiiotid,0.5,0,9,0.010850861058034234,61,27668,0.002204713025878271,9,0
+xiiotid,0.5,1,0,0.012069615819817858,10,121896,8.203714641989892e-05,9,0
+xiiotid,0.5,1,1,-0.0016588430239991014,98,37145,0.002638309328307982,8,1
+xiiotid,0.5,1,2,0.04090973661622571,310,68062,0.004554670741382856,8,1
+xiiotid,0.5,1,3,-0.009330628803245422,177,91761,0.0019289240527021283,9,0
+xiiotid,0.5,1,4,0.023390929820962802,288,51395,0.0056036579433797066,9,0
+xiiotid,0.5,1,5,0.008557414694983834,14,16884,0.0008291873963515755,9,0
+xiiotid,0.5,1,6,-0.0020582497545886813,960,73258,0.01310437085369516,7,2
+xiiotid,0.5,1,7,-0.0002984401302459916,54,52612,0.001026381814034821,9,0
+xiiotid,0.5,1,8,0.06112461987079512,120,29506,0.004066969429946452,9,0
+xiiotid,0.5,1,9,0.012503638392631244,174,27390,0.006352683461117196,9,0
+xiiotid,0.5,2,0,0.3235930735930736,128,27275,0.004692942254812099,9,0
+xiiotid,0.5,2,1,0.028909818130736897,12,10091,0.0011891784758695867,9,0
+xiiotid,0.5,2,2,0.08047542639805078,375,88593,0.004232840066370932,8,1
+xiiotid,0.5,2,3,0.09486679291488886,74,66112,0.0011193126815101646,9,0
+xiiotid,0.5,2,4,0.007974082073434086,65,36350,0.0017881705639614855,9,0
+xiiotid,0.5,2,5,0.030823470158650213,277,29323,0.009446509565869794,9,0
+xiiotid,0.5,2,6,0.009820871011708476,21,87876,0.00023897309845691657,9,0
+xiiotid,0.5,2,7,-0.01395374602845062,39,86394,0.0004514202375164942,9,0
+xiiotid,0.5,2,8,0.14776984457049847,45,99807,0.0004508701794463314,9,0
+xiiotid,0.5,2,9,0.048939405654784884,87,38088,0.002284183994959042,9,0
+xiiotid,0.75,0,0,0.012100380853047654,181,53233,0.0034001465256513815,9,0
+xiiotid,0.75,0,1,0.028491022660640075,231,24236,0.009531275788083842,9,0
+xiiotid,0.75,0,2,0.024781494427492667,156,100573,0.0015511121275093712,8,1
+xiiotid,0.75,0,3,0.010207689906927386,89,29634,0.0030033070122156983,9,0
+xiiotid,0.75,0,4,-0.01697103222518126,202,70610,0.002860784591417646,8,1
+xiiotid,0.75,0,5,0.03315611959069342,61,47117,0.0012946494895685209,9,0
+xiiotid,0.75,0,6,0.015696244925298752,290,28804,0.010068046104707679,9,0
+xiiotid,0.75,0,7,0.06215375530760542,132,54245,0.0024334040003686974,9,0
+xiiotid,0.75,0,8,-0.0004278344531727825,79,82505,0.0009575177261984122,9,0
+xiiotid,0.75,0,9,-0.0024223335938818313,1031,78952,0.013058567230722464,8,1
+xiiotid,0.75,1,0,0.011504273346864324,110,34279,0.003208961755010356,9,0
+xiiotid,0.75,1,1,-0.0034049601987760636,121,95788,0.0012632062471290768,9,0
+xiiotid,0.75,1,2,0.010193569038855876,426,29917,0.014239395661329679,9,0
+xiiotid,0.75,1,3,0.008220247349652376,91,41693,0.002182620583791044,9,0
+xiiotid,0.75,1,4,0.013862245107827853,355,109690,0.003236393472513447,9,0
+xiiotid,0.75,1,5,0.0026610784007268795,462,21623,0.02136613790870832,9,0
+xiiotid,0.75,1,6,0.021536095507448505,36,45425,0.0007925151348376445,9,0
+xiiotid,0.75,1,7,0.005654755072251896,211,42929,0.004915092361806704,9,0
+xiiotid,0.75,1,8,0.016895280781532185,372,55524,0.006699805489518046,8,1
+xiiotid,0.75,1,9,0.06271042675400224,198,93041,0.0021280940660568996,9,0
+xiiotid,0.75,2,0,0.013841721208623592,76,30607,0.002483092103113667,9,0
+xiiotid,0.75,2,1,0.0016817436181817032,82,40979,0.0020010249152004685,9,0
+xiiotid,0.75,2,2,0.004087589009856285,170,42684,0.003982757004966732,9,0
+xiiotid,0.75,2,3,-0.03174244107752611,148,85848,0.001723977262137732,9,0
+xiiotid,0.75,2,4,0.010332069101179266,44,99012,0.0004443905789197269,9,0
+xiiotid,0.75,2,5,-0.0697526762643042,28,39573,0.000707553129659111,9,0
+xiiotid,0.75,2,6,0.005704637240450006,178,24224,0.007348084544253633,9,0
+xiiotid,0.75,2,7,-0.008757536841214941,12,28149,0.00042630288820206754,9,0
+xiiotid,0.75,2,8,0.06759906759906764,87,62937,0.001382334715668049,9,0
+xiiotid,0.75,2,9,0.04065807684333811,101,115896,0.0008714709739766687,9,0
+5gnidd,0.5,0,0,0.01892463962702573,7,116785,5.993920452112857e-05,9,0
+5gnidd,0.5,0,1,0.04316806472122747,132,78161,0.0016888217909187446,9,0
+5gnidd,0.5,0,2,0.06217913201304848,129,98019,0.0013160713739173017,9,0
+5gnidd,0.5,0,3,-0.05739825972017648,496,48536,0.01021921872424592,8,1
+5gnidd,0.5,0,4,0.08578118690523406,19,54279,0.0003500432948285709,8,1
+5gnidd,0.5,0,5,0.034721017618076555,13,126582,0.00010270022594049707,9,0
+5gnidd,0.5,0,6,-0.0003829071592440503,19,43983,0.0004319850851465339,9,0
+5gnidd,0.5,0,7,-0.0905634374183798,94,15388,0.006108656095658955,9,0
+5gnidd,0.5,0,8,-0.007239478035328428,651,178993,0.0036370137379674065,9,0
+5gnidd,0.5,0,9,0.0,735,90397,0.008130800800911536,9,0
+5gnidd,0.5,1,0,-0.17567567567567566,261,131086,0.0019910593045786734,9,0
+5gnidd,0.5,1,1,-0.14463914463914462,181,140144,0.0012915287133234388,9,0
+5gnidd,0.5,1,2,0.016716814651180023,640,31613,0.020244835985195962,9,0
+5gnidd,0.5,1,3,0.10340027838536492,25,57521,0.00043462387649727925,9,0
+5gnidd,0.5,1,4,0.03988304093567253,172,62253,0.0027629190561097457,9,0
+5gnidd,0.5,1,5,0.02840909090909094,105,116940,0.0008978963571062083,9,0
+5gnidd,0.5,1,6,-0.01659848042080636,1203,97485,0.012340360055393138,9,0
+5gnidd,0.5,1,7,0.0,439,124812,0.003517290004166266,9,0
+5gnidd,0.5,1,8,0.022689768976897673,29,53759,0.0005394445581205008,9,0
+5gnidd,0.5,1,9,-0.04018251398292616,393,35510,0.011067304984511405,9,0
+5gnidd,0.5,2,0,0.25892857142857145,25,255463,9.786152984972383e-05,8,1
+5gnidd,0.5,2,1,0.29673733296356775,5,93960,5.321413367390379e-05,9,0
+5gnidd,0.5,2,2,-0.05983168569271902,10,20087,0.0004978344202718176,9,0
+5gnidd,0.5,2,3,0.4040650406504065,80,40162,0.0019919326726756635,9,0
+5gnidd,0.5,2,4,0.02157232998676517,80,99599,0.0008032209158726493,9,0
+5gnidd,0.5,2,5,0.14721254355400692,127,73322,0.0017320858678159352,9,0
+5gnidd,0.5,2,6,0.09375,149,21124,0.007053588335542511,8,1
+5gnidd,0.5,2,7,0.0014146505726297942,89,52714,0.0016883560344500511,9,0
+5gnidd,0.5,2,8,0.0040000000000000036,52,160995,0.00032299139724836177,9,0
+5gnidd,0.5,2,9,0.07997828457131029,154,33697,0.004570139775054159,9,0
+5gnidd,0.75,0,0,0.002491522924959577,254,77880,0.003261427837699024,8,1
+5gnidd,0.75,0,1,4.7653203813946554e-05,108,97748,0.0011048819413184924,9,0
+5gnidd,0.75,0,2,8.599931200548028e-05,130,143547,0.0009056267285279386,9,0
+5gnidd,0.75,0,3,0.04620689062570027,1060,86837,0.012206778216658798,8,1
+5gnidd,0.75,0,4,0.010346119107374596,1980,42381,0.046719048630282436,8,1
+5gnidd,0.75,0,5,-0.03633656881678371,323,49484,0.006527362379759114,9,0
+5gnidd,0.75,0,6,-0.01303298471440062,482,93686,0.005144845547894029,9,0
+5gnidd,0.75,0,7,-0.0014829594477998587,201,38839,0.005175210484307011,9,0
+5gnidd,0.75,0,8,-0.01975396169912713,187,80087,0.002334960730205901,9,0
+5gnidd,0.75,0,9,0.01834553583876719,77,140634,0.000547520514242644,9,0
+5gnidd,0.75,1,0,0.16839803756061722,603,36687,0.016436339847902527,9,0
+5gnidd,0.75,1,1,0.014937839937839925,762,46522,0.016379347405528568,9,0
+5gnidd,0.75,1,2,0.00025068930142924106,163,45258,0.00360157320252773,9,0
+5gnidd,0.75,1,3,0.0010304374919096748,743,202829,0.003663184258661237,8,1
+5gnidd,0.75,1,4,0.037012987012987,72,17861,0.004031129276076367,9,0
+5gnidd,0.75,1,5,0.061520981697536836,867,115584,0.007501038205980066,9,0
+5gnidd,0.75,1,6,0.13236612884064647,55,52144,0.0010547714022706351,8,1
+5gnidd,0.75,1,7,-0.004699533078648788,77,116644,0.0006601282534892493,9,0
+5gnidd,0.75,1,8,0.008754208754208848,286,127130,0.00224966569653111,9,0
+5gnidd,0.75,1,9,0.0,473,90464,0.005228599221789883,9,0
+5gnidd,0.75,2,0,0.0907530728693951,423,78791,0.0053686334733662475,9,0
+5gnidd,0.75,2,1,0.037878787878787845,570,107810,0.005287079120675262,9,0
+5gnidd,0.75,2,2,0.014085531581619848,140,58432,0.002395947426067908,9,0
+5gnidd,0.75,2,3,0.0,31,65107,0.0004761392784186032,9,0
+5gnidd,0.75,2,4,0.0013628477356519841,151,49537,0.0030482265781133296,9,0
+5gnidd,0.75,2,5,0.0064626797036261285,257,71485,0.003595159823739246,9,0
+5gnidd,0.75,2,6,0.0019637279732351853,436,191908,0.0022719219626070827,9,0
+5gnidd,0.75,2,7,0.0,219,89114,0.0024575263146082546,9,0
+5gnidd,0.75,2,8,0.013833992094861691,92,62795,0.0014650847997452027,9,0
+5gnidd,0.75,2,9,0.028593999970401685,474,76144,0.0062250472788400926,9,0
@@ -0,0 +1,91 @@
+dataset,alpha,seed,selected_beta,metric,original,proposed,delta
+xiiotid,0.25,0,0.0,rare_class_macro_f1,0.6398103625587044,0.6413622282749385,0.0015518657162341132
+xiiotid,0.25,0,0.0,rare_class_recall,0.8296012943628512,0.830700667573069,0.0010993732102178022
+xiiotid,0.25,0,0.0,macro_f1,0.6731025782780014,0.6745583013411162,0.001455723063114811
+xiiotid,0.25,0,0.0,weighted_f1,0.9054655196913645,0.9061557688476052,0.0006902491562407187
+xiiotid,0.25,0,0.0,accuracy,0.8737218342944525,0.8745509147521049,0.000829080457652398
+xiiotid,0.25,1,0.0,rare_class_macro_f1,0.5825215614975117,0.5805553678969027,-0.001966193600609012
+xiiotid,0.25,1,0.0,rare_class_recall,0.8266995877156649,0.8256994230374527,-0.0010001646782121787
+xiiotid,0.25,1,0.0,macro_f1,0.6669813754005415,0.6650385577782836,-0.0019428176222578974
+xiiotid,0.25,1,0.0,weighted_f1,0.9020747305215218,0.899863925003673,-0.0022108055178488017
+xiiotid,0.25,1,0.0,accuracy,0.8673409854388906,0.8634903673133494,-0.0038506181255412164
+xiiotid,0.25,2,0.0,rare_class_macro_f1,0.6760389917828752,0.6744891962601014,-0.0015497955227737892
+xiiotid,0.25,2,0.0,rare_class_recall,0.8095946724338723,0.807370314447251,-0.0022243579866212393
+xiiotid,0.25,2,0.0,macro_f1,0.7059686496708458,0.7044332575239617,-0.0015353921468841403
+xiiotid,0.25,2,0.0,weighted_f1,0.9154163421325315,0.9146272396181951,-0.0007891025143363928
+xiiotid,0.25,2,0.0,accuracy,0.8852306993140127,0.8840024319693425,-0.0012282673446701864
+xiiotid,0.5,0,0.75,rare_class_macro_f1,0.6096198579015474,0.6361887569975536,0.026568899096006238
+xiiotid,0.5,0,0.75,rare_class_recall,0.8714872155034664,0.8575434903523123,-0.013943725151154163
+xiiotid,0.5,0,0.75,macro_f1,0.7077011542142579,0.7262984173916737,0.018597263177415724
+xiiotid,0.5,0,0.75,weighted_f1,0.9175853183547894,0.9222775427473946,0.004692224392605193
+xiiotid,0.5,0,0.75,accuracy,0.8904324115186911,0.8886698478790893,-0.0017625636396018107
+xiiotid,0.5,1,0.75,rare_class_macro_f1,0.655665402810942,0.6719373483509694,0.01627194554002742
+xiiotid,0.5,1,0.75,rare_class_recall,0.8893650106408222,0.8518834947182458,-0.037481515922576425
+xiiotid,0.5,1,0.75,macro_f1,0.7091760460445067,0.7290145924987934,0.019838546454286732
+xiiotid,0.5,1,0.75,weighted_f1,0.929132450103975,0.9447335103607959,0.015601060256820842
+xiiotid,0.5,1,0.75,accuracy,0.9008849666218349,0.9254933028723031,0.02460833625046821
+xiiotid,0.5,2,0.75,rare_class_macro_f1,0.662274505162012,0.6883814197743454,0.026106914612333387
+xiiotid,0.5,2,0.75,rare_class_recall,0.8700179539413673,0.878321192150252,0.008303238208884678
+xiiotid,0.5,2,0.75,macro_f1,0.7100168446133323,0.7381943369820925,0.028177492368760215
+xiiotid,0.5,2,0.75,weighted_f1,0.9274806689914715,0.9338838836432976,0.006403214651826028
+xiiotid,0.5,2,0.75,accuracy,0.9070447273553561,0.9050119448999269,-0.0020327824554292606
+xiiotid,0.75,0,0.5,rare_class_macro_f1,0.5215401253860467,0.5118888965076885,-0.00965122887835812
+xiiotid,0.75,0,0.5,rare_class_recall,0.8164397897152784,0.8038871133887995,-0.01255267632647894
+xiiotid,0.75,0,0.5,macro_f1,0.7424713724903063,0.741976732126316,-0.000494640363990273
+xiiotid,0.75,0,0.5,weighted_f1,0.9325633790408211,0.9355656459425888,0.0030022669017677472
+xiiotid,0.75,0,0.5,accuracy,0.9132536187826642,0.9143652007295908,0.0011115819469266208
+xiiotid,0.75,1,0.5,rare_class_macro_f1,0.6140607772418322,0.6241440493793726,0.010083272137540455
+xiiotid,0.75,1,0.5,rare_class_recall,0.8832376049795277,0.8767262170299326,-0.00651138794959516
+xiiotid,0.75,1,0.5,macro_f1,0.7473954752926397,0.7617524315559205,0.014356956263280818
+xiiotid,0.75,1,0.5,weighted_f1,0.9359372345770101,0.9483435473557725,0.012406312778762474
+xiiotid,0.75,1,0.5,accuracy,0.9138984591386161,0.9326111121346672,0.01871265299605107
+xiiotid,0.75,2,0.5,rare_class_macro_f1,0.6760680764285393,0.6797040888050662,0.003636012376526887
+xiiotid,0.75,2,0.5,rare_class_recall,0.8698349315628103,0.8656806653880937,-0.004154266174716592
+xiiotid,0.75,2,0.5,macro_f1,0.7656800124975965,0.7691786586117774,0.0034986461141808256
+xiiotid,0.75,2,0.5,weighted_f1,0.9489075724365953,0.9517397334421744,0.002832161005579148
+xiiotid,0.75,2,0.5,accuracy,0.9345763398861396,0.9373522240850943,0.0027758841989546923
+5gnidd,0.25,0,0.0,rare_class_macro_f1,0.8649030079014913,0.8598927442995293,-0.005010263601961995
+5gnidd,0.25,0,0.0,rare_class_recall,0.9854440969271964,0.9852344223763528,-0.00020967455084353048
+5gnidd,0.25,0,0.0,macro_f1,0.8350430568226094,0.8296799384783186,-0.005363118344290818
+5gnidd,0.25,0,0.0,weighted_f1,0.7180410173151626,0.7174035576457394,-0.0006374596694231904
+5gnidd,0.25,0,0.0,accuracy,0.745280186859777,0.7448442903728559,-0.0004358964869211057
+5gnidd,0.25,1,0.0,rare_class_macro_f1,0.8148479924290983,0.8138885616043564,-0.0009594308247419248
+5gnidd,0.25,1,0.0,rare_class_recall,0.9723332887819381,0.9726587864975359,0.00032549771559775564
+5gnidd,0.25,1,0.0,macro_f1,0.8083728869633723,0.8029871664481885,-0.005385720515183823
+5gnidd,0.25,1,0.0,weighted_f1,0.7163267008884987,0.7123024521167044,-0.004024248771794303
+5gnidd,0.25,1,0.0,accuracy,0.7412954350123573,0.737779477499928,-0.0035159575124292797
+5gnidd,0.25,2,0.0,rare_class_macro_f1,0.8470722146704726,0.8499254383428447,0.0028532236723721738
+5gnidd,0.25,2,0.0,rare_class_recall,0.9238897222901731,0.9257555179886866,0.0018657956985135282
+5gnidd,0.25,2,0.0,macro_f1,0.8148003580492102,0.8172541498613525,0.0024537918121422253
+5gnidd,0.25,2,0.0,weighted_f1,0.6819385320735543,0.6836842063399526,0.0017456742663983782
+5gnidd,0.25,2,0.0,accuracy,0.7063291347454734,0.7080521595381142,0.0017230247926407483
+5gnidd,0.5,0,1.0,rare_class_macro_f1,0.8719637151922778,0.8809620350206805,0.008998319828402712
+5gnidd,0.5,0,1.0,rare_class_recall,0.9743136055202825,0.9743551890147765,4.1583494494012996e-05
+5gnidd,0.5,0,1.0,macro_f1,0.8368808854359121,0.8504696755820977,0.013588790146185636
+5gnidd,0.5,0,1.0,weighted_f1,0.7087651452654392,0.7169500389469496,0.00818489368151043
+5gnidd,0.5,0,1.0,accuracy,0.7344156725348203,0.7428375216406157,0.008421849105795354
+5gnidd,0.5,1,1.0,rare_class_macro_f1,0.9211362509007983,0.8400749080738759,-0.08106134282692246
+5gnidd,0.5,1,1.0,rare_class_recall,0.9685663953100676,0.9683408980660666,-0.00022549724400100501
+5gnidd,0.5,1,1.0,macro_f1,0.8654374620417004,0.8159411532871907,-0.04949630875450961
+5gnidd,0.5,1,1.0,weighted_f1,0.7228023713469742,0.7201060072251736,-0.002696364121800632
+5gnidd,0.5,1,1.0,accuracy,0.7473774246742085,0.7440259564021269,-0.003351468272081659
+5gnidd,0.5,2,1.0,rare_class_macro_f1,0.8485250688106042,0.8949144498720359,0.046389381061431756
+5gnidd,0.5,2,1.0,rare_class_recall,0.9758591365273065,0.9586158518509788,-0.01724328467632774
+5gnidd,0.5,2,1.0,macro_f1,0.8248673149789701,0.8493777051701392,0.024510390191169074
+5gnidd,0.5,2,1.0,weighted_f1,0.7246221247269671,0.7175295566399255,-0.0070925680870416175
+5gnidd,0.5,2,1.0,accuracy,0.7501326194500302,0.7429074295677633,-0.007225189882266836
+5gnidd,0.75,0,0.5,rare_class_macro_f1,0.9055885946257963,0.9044472051910231,-0.0011413894347732656
+5gnidd,0.75,0,0.5,rare_class_recall,0.9612851237375019,0.9586662658643216,-0.0026188578731802803
+5gnidd,0.75,0,0.5,macro_f1,0.8571092007836155,0.8573089581866151,0.00019975740299960076
+5gnidd,0.75,0,0.5,weighted_f1,0.7210591954237406,0.7238852910350473,0.002826095611306645
+5gnidd,0.75,0,0.5,accuracy,0.7470484461935134,0.7495075603367095,0.002459114143196106
+5gnidd,0.75,1,0.5,rare_class_macro_f1,0.8607580175603599,0.9057207431633723,0.0449627256030124
+5gnidd,0.75,1,0.5,rare_class_recall,0.9590357221091453,0.9597228673347585,0.0006871452256131949
+5gnidd,0.75,1,0.5,macro_f1,0.8298666656506455,0.8599385529838828,0.03007188733323729
+5gnidd,0.75,1,0.5,weighted_f1,0.7299504931440421,0.7301094421086752,0.00015894896463308772
+5gnidd,0.75,1,0.5,accuracy,0.7540721367563544,0.7543805540820061,0.00030841732565167757
+5gnidd,0.75,2,0.5,rare_class_macro_f1,0.8733557623287002,0.8800449253698305,0.006689163041130319
+5gnidd,0.75,2,0.5,rare_class_recall,0.9633214115686995,0.960715099133502,-0.0026063124351975997
+5gnidd,0.75,2,0.5,macro_f1,0.8372382392225877,0.842201195406295,0.0049629561837073455
+5gnidd,0.75,2,0.5,weighted_f1,0.7271326093968319,0.7284075566369677,0.0012749472401357709
+5gnidd,0.75,2,0.5,accuracy,0.7517446140054364,0.7529330487669476,0.0011884347615112034
@@ -0,0 +1,31 @@
+dataset,alpha,selected_beta,metric,orig_mean,orig_std,prop_mean,prop_std,delta_mean,delta_std,per_seed_deltas,n
+xiiotid,0.25,0.0,rare_class_macro_f1,0.632790305279697,0.04715228886220545,0.6321355974773142,0.047641778563026725,-0.0006547078023828959,0.0019222569702312443,0.001552|-0.001966|-0.001550,3
+xiiotid,0.25,0.0,rare_class_recall,0.8219651848374628,0.010810973971484781,0.8212568016859243,0.012283281629867223,-0.0007083831515385386,0.0016809668681122992,0.001099|-0.001000|-0.002224,3
+xiiotid,0.25,0.0,macro_f1,0.6820175344497962,0.02096686007526524,0.6813433722144538,0.020555132742947534,-0.0006741622353424089,0.0018557498263327506,0.001456|-0.001943|-0.001535,3
+xiiotid,0.25,0.0,weighted_f1,0.9076521974484727,0.0069343939753111275,0.9068823111564911,0.007408425095578143,-0.0007698862919814919,0.0014506227982888447,0.000690|-0.002211|-0.000789,3
+xiiotid,0.25,0.0,accuracy,0.8754311730157852,0.009066523301774463,0.8740145713449322,0.010266545051804106,-0.0014166016708530016,0.00234552703484742,0.000829|-0.003851|-0.001228,3
+xiiotid,0.5,0.75,rare_class_macro_f1,0.6425199219581671,0.028683282930463034,0.6655025083742895,0.026684712591240056,0.022982586416122348,0.005816174270233457,0.026569|0.016272|0.026107,3
+xiiotid,0.5,0.75,rare_class_recall,0.8769567266952186,0.010770970951613777,0.86258272574027,0.013920609948849614,-0.014374000954948637,0.022895409598216065,-0.013944|-0.037482|0.008303,3
+xiiotid,0.5,0.75,macro_f1,0.7089646816240323,0.0011722251041387443,0.7311691156241865,0.006233755140268,0.022204434000154222,0.005209919936951091,0.018597|0.019839|0.028177,3
+xiiotid,0.5,0.75,weighted_f1,0.9247328124834121,0.006244765764216191,0.9336316455838293,0.011230108565482132,0.008898833100417355,0.005867005927967981,0.004692|0.015601|0.006403,3
+xiiotid,0.5,0.75,accuracy,0.8994540351652941,0.008398091032446977,0.9063916985504398,0.018450460685614026,0.006937663385145713,0.015303848018947405,-0.001763|0.024608|-0.002033,3
+xiiotid,0.75,0.5,rare_class_macro_f1,0.6038896596854727,0.07776445613264914,0.6052456782307091,0.08548886486265879,0.0013560185452364075,0.010062872924253166,-0.009651|0.010083|0.003636,3
+xiiotid,0.75,0.5,rare_class_recall,0.8565041087525388,0.03533794219558259,0.8487646652689419,0.039255535276369505,-0.007739443483596897,0.004331791034347159,-0.012553|-0.006511|-0.004154,3
+xiiotid,0.75,0.5,macro_f1,0.7518489534268475,0.012228464472937675,0.7576359407646712,0.014060416901925107,0.00578698733782379,0.007685691885683445,-0.000495|0.014357|0.003499,3
+xiiotid,0.75,0.5,weighted_f1,0.9391360620181421,0.008628878095570825,0.9452163089135119,0.008528481213975155,0.00608024689536979,0.005479193934763414,0.003002|0.012406|0.002832,3
+xiiotid,0.75,0.5,accuracy,0.92057613926914,0.012128815595858657,0.9281095123164508,0.012136684529934048,0.0075333730473107945,0.009717237293889465,0.001112|0.018713|0.002776,3
+5gnidd,0.25,0.0,rare_class_macro_f1,0.842274405000354,0.02537006851997817,0.8412355814155769,0.024201888695409735,-0.0010388235847772487,0.003932344775851092,-0.005010|-0.000959|0.002853,3
+5gnidd,0.25,0.0,rare_class_recall,0.9605557026664359,0.032423277290702865,0.9612162422875251,0.03134699156678076,0.0006605396210892511,0.001077536077269848,-0.000210|0.000325|0.001866,3
+5gnidd,0.25,0.0,macro_f1,0.8194054339450639,0.013918675694370832,0.8166404182626198,0.01335696520619808,-0.0027650156824441385,0.004519633996652314,-0.005363|-0.005386|0.002454,3
+5gnidd,0.25,0.0,weighted_f1,0.7054354167590718,0.02036694411653514,0.7044634053674654,0.018175165847202895,-0.0009720113916063718,0.0028994735052727187,-0.000637|-0.004024|0.001746,3
+5gnidd,0.25,0.0,accuracy,0.7309682522058693,0.021430915606706246,0.730225309136966,0.019524710435296355,-0.0007429430689032124,0.0026329531140602156,-0.000436|-0.003516|0.001723,3
+5gnidd,0.5,1.0,rare_class_macro_f1,0.8805416783012268,0.0370578200850738,0.8719837976555308,0.028500884123071776,-0.008557880645695995,0.06551402053172613,0.008998|-0.081061|0.046389,3
+5gnidd,0.5,1.0,rare_class_recall,0.9729130457858854,0.0038428107281798614,0.9671039796439406,0.007942238914573898,-0.0058090661419449114,0.009903224129001456,0.000042|-0.000225|-0.017243,3
+5gnidd,0.5,1.0,macro_f1,0.8423952208188609,0.02083962881650294,0.8385961780131425,0.019627422368850796,-0.0037990428057183,0.039949974008521794,0.013589|-0.049496|0.024510,3
+5gnidd,0.5,1.0,weighted_f1,0.7187298804464602,0.00867754781098361,0.7181952009373496,0.0016799838366213199,-0.0005346795091106064,0.007864786697422122,0.008185|-0.002696|-0.007093,3
+5gnidd,0.5,1.0,accuracy,0.7439752388863531,0.008392660852506539,0.743256969203502,0.0006668791219600918,-0.0007182696828510471,0.008149095556025167,0.008422|-0.003351|-0.007225,3
+5gnidd,0.75,0.5,rare_class_macro_f1,0.879900791504952,0.023120839413686988,0.8967376245747419,0.014470318925337908,0.01683683306978982,0.024670402532432648,-0.001141|0.044963|0.006689,3
+5gnidd,0.75,0.5,rare_class_recall,0.9612140858051156,0.002143727670860224,0.9597014107775274,0.0010245851497612311,-0.0015126750275882284,0.0019051105497581826,-0.002619|0.000687|-0.002606,3
+5gnidd,0.75,0.5,macro_f1,0.8414047018856162,0.014091078328488028,0.8531495688589309,0.009572296127013734,0.011744866973314746,0.01604935425837736,0.000200|0.030072|0.004963,3
+5gnidd,0.75,0.5,weighted_f1,0.7260474326548715,0.004543897028843018,0.7274674299268967,0.0032168140791904,0.001419997272025168,0.0013394765554233625,0.002826|0.000159|0.001275,3
+5gnidd,0.75,0.5,accuracy,0.7509550656517682,0.0035777922288362054,0.7522737210618877,0.0025025090963709863,0.0013186554101196624,0.0010812456997565073,0.002459|0.000308|0.001188,3
@@ -0,0 +1,163 @@
+dataset,alpha,seed,class_id,class_name,support,orig_precision,prop_precision,delta_precision,orig_recall,prop_recall,delta_recall,orig_f1,prop_f1,delta_f1
+xiiotid,0.25,0,0,C&C,573,0.07959847306659126,0.08086756679115197,0.001269093724560702,0.9825479930191972,0.9825479930191972,0.0,0.14726654459848285,0.14943596549435964,0.0021694208958767947
+xiiotid,0.25,0,1,Exfiltration,4427,1.0,1.0,0.0,0.960018070928394,0.960018070928394,0.0,0.9796012446698167,0.9796012446698167,0.0
+xiiotid,0.25,0,2,Exploitation,227,0.016145307769929364,0.016145181476846057,-1.2629308330722178e-07,0.5638766519823789,0.5682819383259912,0.004405286343612258,0.03139178418148375,0.03139832055494706,6.536373463310319e-06
+xiiotid,0.25,0,3,Lateral_movement,6319,0.9509666573269824,0.9515994436717663,0.0006327863447839199,0.5371103022630163,0.5413831302421269,0.0042728279791106605,0.6864886731391586,0.6901351623966108,0.0036464892574521457
+xiiotid,0.25,0,4,Normal,83651,0.9706643548325408,0.9707035726108324,3.921777829163364e-05,0.9101624606998123,0.9114176758197751,0.0012552151199627382,0.9394403040323774,0.940126885870537,0.0006865818381596167
+xiiotid,0.25,0,5,RDOS,28252,0.9924600098654076,0.9923900789177001,-6.993094770746389e-05,0.9970267591674926,0.9970267591674926,0.0,0.9947381431648833,0.9947030157496999,-3.512741518341134e-05
+xiiotid,0.25,0,6,Reconnaissance,25474,0.9939514871858827,0.99401384298809,6.235580220737003e-05,0.6257360445944885,0.6257753003061945,3.925571170604325e-05,0.7679892076798921,0.7680373876803739,4.818000048179005e-05
+xiiotid,0.25,0,7,Tampering,1024,0.42526315789473684,0.42997020008514264,0.004707042190405797,0.986328125,0.986328125,0.0,0.5942924389526332,0.5988734064630893,0.004580967510456113
+xiiotid,0.25,0,8,Weaponization,12884,0.857916102841678,0.8613639451161527,0.0034478422744746995,0.9841664079478423,0.9842440235951568,7.761564731456616e-05,0.9167148640832852,0.9187133231906106,0.0019984591073254343
+xiiotid,0.25,1,0,C&C,573,0.06773003213903107,0.06334892006234692,-0.004381112076684152,0.9930191972076788,0.9930191972076788,0.0,0.126810786717183,0.11909994767137626,-0.0077108390458067305
+xiiotid,0.25,1,1,Exfiltration,4427,0.9995480225988701,0.9995480225988701,0.0,0.9990964535803026,0.9990964535803026,0.0,0.999322187076367,0.999322187076367,0.0
+xiiotid,0.25,1,2,Exploitation,227,0.01957343412526998,0.019405781584582442,-0.00016765254068753718,0.6387665198237885,0.6387665198237885,0.0,0.03798297314996725,0.03766722951032602,-0.00031574363964122915
+xiiotid,0.25,1,3,Lateral_movement,6319,0.9467521605798718,0.9461060948081265,-0.0006460657717453078,0.5374268080392467,0.5306219338502928,-0.006804874188953924,0.6856450635978194,0.6799148332150461,-0.005730230382773316
+xiiotid,0.25,1,4,Normal,83651,0.9718254744668362,0.9717519322782481,-7.354218858812711e-05,0.8906647858363916,0.8837551254617398,-0.006909660374651794,0.9294767833528781,0.9256669191808524,-0.0038098641720256676
+xiiotid,0.25,1,5,RDOS,28252,0.9944917199251438,0.9947022674295402,0.00021054750439630787,0.996920571994903,0.9968851762707065,-3.53957241965297e-05,0.995704664769413,0.9957925255453806,8.786077596756314e-05
+xiiotid,0.25,1,6,Reconnaissance,25474,0.9575149824956981,0.9568208778173191,-0.0006941046783790483,0.633469419800581,0.6332731412420507,-0.00019627855853032727,0.7624920263661493,0.7621297302404686,-0.0003622961256807278
+xiiotid,0.25,1,7,Tampering,1024,0.35095652173913044,0.3508344923504868,-0.00012202938864364077,0.9853515625,0.9853515625,0.0,0.5175686073352141,0.5174358974358975,-0.0001327098993166853
+xiiotid,0.25,1,8,Weaponization,12884,0.9010212646894236,0.9019044951687438,0.0008832304793202628,0.9997671530580565,0.9997671530580565,0.0,0.9478292862398823,0.9483177501288375,0.0004884638889551818
+xiiotid,0.25,2,0,C&C,573,0.09334763948497854,0.09018659295093297,-0.0031610465340455723,0.9109947643979057,0.9109947643979057,0.0,0.16934306569343063,0.1641251375569879,-0.005217928136442729
+xiiotid,0.25,2,1,Exfiltration,4427,0.998820406959599,0.9987933634992459,-2.704346035309335e-05,0.7650779308786989,0.74791054890445,-0.01716738197424894,0.8664620107444359,0.8553345388788427,-0.011127471865593153
+xiiotid,0.25,2,2,Exploitation,227,0.016488046166529265,0.016525977886101533,3.793171957226782e-05,0.6167400881057269,0.6123348017621145,-0.004405286343612369,0.03211745813259922,0.032183375781430885,6.591764883166762e-05
+xiiotid,0.25,2,3,Lateral_movement,6319,0.9145810663764962,0.9127383951061533,-0.0018426712703428771,0.7980693147649944,0.8028169014084507,0.004747586643456314,0.852362038367278,0.8542561252841626,0.0018940869168846408
+xiiotid,0.25,2,4,Normal,83651,0.9710152233088013,0.9709588140883874,-5.640922041394436e-05,0.9211127183177726,0.918470789350994,-0.002641928966778617,0.9454059127746115,0.9439857476348446,-0.001420165139766949
+xiiotid,0.25,2,5,RDOS,28252,0.9919369036301539,0.9920766278127972,0.0001397241826432971,0.9971683420642786,0.9971683420642786,0.0,0.9945457433851694,0.9946159685078289,7.02251226594175e-05
+xiiotid,0.25,2,6,Reconnaissance,25474,0.9916464891041162,0.9916451283284246,-1.360775691683358e-06,0.6430870691685641,0.6476407317264662,0.004553662557902127,0.7802066961946944,0.7835482415520886,0.003341545357394171
+xiiotid,0.25,2,7,Tampering,1024,0.8247096092925026,0.8340471092077087,0.00933749991520616,0.7626953125,0.7607421875,-0.001953125,0.7924911212582446,0.7957099080694586,0.003218786811214014
+xiiotid,0.25,2,8,Weaponization,12884,0.8660329617725501,0.8605920065475379,-0.005440955225012201,0.9829245575908103,0.9793542378143434,-0.0035703197764669348,0.9207838004871487,0.9161402744500109,-0.004643526037137802
+xiiotid,0.5,0,0,C&C,573,0.09224011713030747,0.08453441295546558,-0.007705704174841885,0.9895287958115183,0.9109947643979057,-0.07853403141361259,0.16875,0.15471250740960285,-0.01403749259039716
+xiiotid,0.5,0,1,Exfiltration,4427,1.0,1.0,0.0,0.9516602665461938,0.9602439575333183,0.008583690987124415,0.9752314814814814,0.9797188292233234,0.004487347741842007
+xiiotid,0.5,0,2,Exploitation,227,0.018892228424216402,0.014540647719762063,-0.004351580704454339,0.5814977973568282,0.5814977973568282,0.0,0.03659550873301912,0.028371843095110157,-0.008223665637908967
+xiiotid,0.5,0,3,Lateral_movement,6319,0.9643082091119043,0.9651458782063834,0.0008376690944791321,0.7268555151131508,0.7800284855198607,0.053172970406709874,0.8289117487818085,0.8627691230526868,0.033857374270878315
+xiiotid,0.5,0,4,Normal,83651,0.9706242809264168,0.9705492305767547,-7.505034966215351e-05,0.9278430622467155,0.9273768394878722,-0.00046622275884333764,0.9487516425755586,0.9484720108081011,-0.00027963176745748886
+xiiotid,0.5,0,5,RDOS,28252,0.9924255768892021,0.9925644007470839,0.00013882385788177576,0.9970975506158856,0.9969559677190996,-0.00014158289678600777,0.9947560781821071,0.9947553373713115,-7.408107955964027e-07
+xiiotid,0.5,0,6,Reconnaissance,25474,0.9917997142324657,0.9944206008583691,0.002620886625903429,0.6267174373871398,0.6366883881604773,0.009970950773337539,0.7680835197613721,0.7763258663603292,0.008242346598957062
+xiiotid,0.5,0,7,Tampering,1024,0.5628808864265928,0.7361918604651163,0.17331097403852347,0.9921875,0.9892578125,-0.0029296875,0.7182750088370449,0.8441666666666666,0.12589165782962175
+xiiotid,0.5,0,8,Weaponization,12884,0.8789910158949551,0.9728469779995093,0.09385596210455416,0.9871934181931077,0.9232381248059609,-0.06395529338714678,0.9299553995759303,0.9473935725379315,0.01743817296200123
+xiiotid,0.5,1,0,C&C,573,0.08977834476160101,0.09979016087666123,0.010011816115060218,0.9825479930191972,0.7469458987783595,-0.23560209424083767,0.16452367036820573,0.1760592348827643,0.01153556451455856
+xiiotid,0.5,1,1,Exfiltration,4427,0.9997687326549491,1.0,0.00023126734505085622,0.9765077930878698,0.9552744522249831,-0.02123334086288675,0.9880013712718546,0.9771256931608133,-0.010875678111041287
+xiiotid,0.5,1,2,Exploitation,227,0.0234779792746114,0.031936579841449604,0.008458600566838205,0.6387665198237885,0.6211453744493393,-0.017621145374449254,0.04529126971732001,0.06074967686342094,0.01545840714610093
+xiiotid,0.5,1,3,Lateral_movement,6319,0.9541181736794987,0.9593312245820154,0.005213050902516669,0.6746320620351322,0.6719417629371736,-0.002690299097958593,0.790395846852693,0.7903210795718939,-7.47672807991373e-05
+xiiotid,0.5,1,4,Normal,83651,0.9725,0.9710573675517097,-0.0014426324482903485,0.9161038122676358,0.9417460640040167,0.025642251736380905,0.9434598740543302,0.9561771363721879,0.012717262317857747
+xiiotid,0.5,1,5,RDOS,28252,0.9948775921150246,0.9965285157633723,0.0016509236483477352,0.9968143848223134,0.995752513096418,-0.0010618717258954469,0.9958450467653247,0.9961403633015828,0.0002953165362581389
+xiiotid,0.5,1,6,Reconnaissance,25474,0.99514972585407,0.992357048499717,-0.002792677354353068,0.7409908141634608,0.8257046400251237,0.08471382586166287,0.8494667206696368,0.9013927576601671,0.05192603699053033
+xiiotid,0.5,1,7,Tampering,1024,0.4755278310940499,0.5771428571428572,0.10161502604880729,0.9677734375,0.986328125,0.0185546875,0.6377091377091377,0.7281903388608507,0.09048120115171299
+xiiotid,0.5,1,8,Weaponization,12884,0.9481128563984218,0.9643887623386485,0.016275905940226654,0.9885128841974542,0.9857963365414467,-0.002716547656007484,0.9678914769920584,0.9749750518154603,0.007083574823401939
+xiiotid,0.5,2,0,C&C,573,0.1745398773006135,0.07111611048618922,-0.10342376681442426,0.9930191972076788,0.9930191972076788,0.0,0.29689538220714845,0.13272684861208303,-0.16416853359506542
+xiiotid,0.5,2,1,Exfiltration,4427,0.9997638724911452,0.9997647058823529,8.333912077107541e-07,0.9564038852496047,0.9597921843234696,0.003388299073864842,0.9776033248672362,0.9793707502593062,0.0017674253920699767
+xiiotid,0.5,2,2,Exploitation,227,0.020633750921149593,0.031478102189781025,0.010844351268631432,0.6167400881057269,0.6079295154185022,-0.008810572687224738,0.0399315459212778,0.05985686402081979,0.01992531809954199
+xiiotid,0.5,2,3,Lateral_movement,6319,0.9250299880047981,0.9301369863013699,0.005106998296571796,0.7322361133090679,0.8596296882418104,0.1273935749327425,0.8174189559226216,0.8934945308002303,0.07607557487760874
+xiiotid,0.5,2,4,Normal,83651,0.9704520181344934,0.9706830191140975,0.00023100097960404575,0.9493610357317904,0.8893856618570011,-0.05997537387478935,0.9597906746271543,0.9282577232120577,-0.0315329514150966
+xiiotid,0.5,2,5,RDOS,28252,0.992565972589226,0.994596312778131,0.002030340188905022,0.9971683420642786,0.9967789890981169,-0.00038935296616171566,0.9948618345546041,0.9956864547608104,0.0008246202062063279
+xiiotid,0.5,2,6,Reconnaissance,25474,0.9787918845910685,0.9713854042670085,-0.007406480324059994,0.6685247703540865,0.8328884352673314,0.16436366491324483,0.7944393907587527,0.8968213712063572,0.10238198044760449
+xiiotid,0.5,2,7,Tampering,1024,0.43615344193378874,0.7971602434077079,0.3610068014739191,0.810546875,0.767578125,-0.04296875,0.567133583874274,0.782089552238806,0.214955968364532
+xiiotid,0.5,2,8,Weaponization,12884,0.9035706649561684,0.9876680722412284,0.08409740728505999,0.9840111766532132,0.9635206457621857,-0.020490530891027592,0.9420769087869217,0.975444937728362,0.03336802894144031
+xiiotid,0.75,0,0,C&C,573,0.17078443483631872,0.14609375,-0.024690684836318727,0.9650959860383944,0.9790575916230366,0.013961605584642212,0.2902125426397271,0.25424881033310676,-0.035963732306620366
+xiiotid,0.75,0,1,Exfiltration,4427,0.9995259540175397,0.9997432605905007,0.00021730657296092915,0.9525638129658911,0.8796024395753331,-0.07296137339055797,0.9754799907471663,0.935832732516222,-0.03964725823094428
+xiiotid,0.75,0,2,Exploitation,227,0.019334993408524976,0.019906499773789776,0.0005715063652648004,0.5814977973568282,0.5814977973568282,0.0,0.037425574142330595,0.03849518810148732,0.0010696139591567233
+xiiotid,0.75,0,3,Lateral_movement,6319,0.9320566194837635,0.9318484042553191,-0.0002082152284443861,0.8857414147808198,0.8871656907738567,0.0014242759930368498,0.9083089905874715,0.9089582488852858,0.0006492582978142591
+xiiotid,0.75,0,4,Normal,83651,0.9711247860758111,0.9714045518787465,0.00027976580293542774,0.956473921411579,0.9531147266619646,-0.003359194749614436,0.9637436762225969,0.9621727298508994,-0.0015709463716975236
+xiiotid,0.75,0,5,RDOS,28252,0.9921110093681764,0.9921110093681764,0.0,0.9970975506158856,0.9970975506158856,0.0,0.9945980298697172,0.9945980298697172,0.0
+xiiotid,0.75,0,6,Reconnaissance,25474,0.983384137230291,0.9943473495991233,0.010963212368832242,0.6458742246996938,0.6767292141006517,0.030854989400957877,0.7796706551356475,0.8053537642194762,0.02568310908382876
+xiiotid,0.75,0,7,Tampering,1024,0.800203873598369,0.8677595628415301,0.06755568924316113,0.7666015625,0.775390625,0.0087890625,0.7830423940149626,0.8189788550799381,0.0359364610649755
+xiiotid,0.75,0,8,Weaponization,12884,0.9104435110699793,0.9297005900779486,0.019257079007969247,0.9926265135051227,0.9905308910276311,-0.00209562247749151,0.9497604990531358,0.9591522302807111,0.009391731227575306
+xiiotid,0.75,1,0,C&C,573,0.10998834045860863,0.1218197498921949,0.011831409433586279,0.987783595113438,0.9860383944153578,-0.0017452006980802626,0.19793670222066795,0.21684897332565722,0.018912271104989276
+xiiotid,0.75,1,1,Exfiltration,4427,0.9997648165569144,1.0,0.0002351834430855959,0.9602439575333183,0.9604698441382427,0.00022588660492439683,0.9796059453854131,0.9798363866804932,0.00023044129508009714
+xiiotid,0.75,1,2,Exploitation,227,0.026335040234089245,0.04698205546492659,0.020647015230837347,0.6343612334801763,0.6343612334801763,0.0,0.050570676031606664,0.08748481166464156,0.036914135633034896
+xiiotid,0.75,1,3,Lateral_movement,6319,0.9445538057742782,0.9406710974225968,-0.003882708351681363,0.9112201297673682,0.9183415097325526,0.00712137996518436,0.9275875956504227,0.929372197309417,0.0017846016589943003
+xiiotid,0.75,1,4,Normal,83651,0.9729585703486139,0.9725495513262992,-0.0004090190223147161,0.9385303224109693,0.9432164588588301,0.004686136447860756,0.9554343989972132,0.957658439485614,0.002224040488400858
+xiiotid,0.75,1,5,RDOS,28252,0.9938247644588729,0.9931953601523111,-0.000629404306561776,0.9968851762707065,0.9970975506158856,0.00021237434517906717,0.9953526179074411,0.9951426300450411,-0.00020998786240000733
+xiiotid,0.75,1,6,Reconnaissance,25474,0.9917502787068004,0.9910376901463036,-0.0007125885604968296,0.698437622674099,0.8030540943707309,0.10461647169663191,0.819643432993965,0.8871975019516003,0.06755406895763527
+xiiotid,0.75,1,7,Tampering,1024,0.857,0.880469583778015,0.023469583778015002,0.8369140625,0.8056640625,-0.03125,0.8468379446640316,0.8414074451810301,-0.005430499483001561
+xiiotid,0.75,1,8,Weaponization,12884,0.9187711346140011,0.9342963995013566,0.015525264887355505,0.9911518162061471,0.9889009624340267,-0.0022508537721204203,0.9535899637829967,0.9608234983597904,0.007233534576793765
+xiiotid,0.75,2,0,C&C,573,0.18441127694859039,0.20878274268104777,0.02437146573245738,0.9703315881326352,0.9458987783595113,-0.024432809773123898,0.30992196209587514,0.34206374250552224,0.0321417804096471
+xiiotid,0.75,2,1,Exfiltration,4427,0.9997643732327992,0.9997643732327992,0.0,0.9584368646939236,0.9584368646939236,0.0,0.9786645138968977,0.9786645138968977,0.0
+xiiotid,0.75,2,2,Exploitation,227,0.03358925143953935,0.03121516164994426,-0.00237408978959509,0.6167400881057269,0.6167400881057269,0.0,0.06370875995449374,0.059422750424448216,-0.004286009530045522
+xiiotid,0.75,2,3,Lateral_movement,6319,0.9369604170060271,0.9413758048119282,0.0044153878059011165,0.910270612438677,0.8792530463680962,-0.031017566070580815,0.9234227002729171,0.9092545618198185,-0.014168138453098611
+xiiotid,0.75,2,4,Normal,83651,0.9713092803583601,0.9713518027667841,4.252240842406163e-05,0.953903719023084,0.9594147111212059,0.005510992098121892,0.9625278191588813,0.9653463560147708,0.0028185368558895085
+xiiotid,0.75,2,5,RDOS,28252,0.9919380390776272,0.9919031190593537,-3.4920018273543185e-05,0.9973099249610647,0.9973099249610647,0.0,0.9946167287360786,0.9945991739913164,-1.755474476228258e-05
+xiiotid,0.75,2,6,Reconnaissance,25474,0.9921287128712871,0.9928634275100205,0.0007347146387334247,0.7867237183010128,0.7973620161733532,0.010638297872340385,0.8775671060121732,0.8844378646695115,0.006870758657338216
+xiiotid,0.75,2,7,Tampering,1024,0.8721311475409836,0.8248031496062992,-0.04732799793468445,0.779296875,0.818359375,0.0390625,0.8231046931407943,0.8215686274509804,-0.0015360656898139036
+xiiotid,0.75,2,8,Weaponization,12884,0.9326123740160377,0.9592397526906343,0.026627378674596636,0.9839335610058988,0.975395839801304,-0.00853772120459484,0.957585829210258,0.9672503367327304,0.009664507522472388
+5gnidd,0.25,0,0,Benign,95547,1.0,1.0,0.0,0.37155536018922625,0.3703517640532931,-0.001203596135933127,0.5418014773212868,0.5405207243399295,-0.0012807529813573248
+5gnidd,0.25,0,1,HTTPFlood,28162,0.9753919782849387,0.9743306750213614,-0.00106130326357734,0.9697464668702507,0.971770470847241,0.0020240039769903184,0.9725610298961913,0.9730488888888889,0.00048785899269754296
+5gnidd,0.25,0,2,ICMPFlood,231,0.6209677419354839,0.5818639798488665,-0.039103762086617366,1.0,1.0,0.0,0.7661691542288557,0.7356687898089173,-0.03050036441993842
+5gnidd,0.25,0,3,SYNFlood,1944,0.7329520457545095,0.7041420118343196,-0.02881003392018988,0.8569958847736625,0.8569958847736625,0.0,0.7901351671804601,0.7730858468677494,-0.01704932031271067
+5gnidd,0.25,0,4,SYNScan,4009,0.9908302354399009,0.9910758552305404,0.00024561979063952855,0.9972561736093789,0.9972561736093789,0.0,0.9940328194927898,0.9941564093000125,0.00012358980722271262
+5gnidd,0.25,0,5,SlowrateDoS,14625,0.9410251201841695,0.9430750374506333,0.002049917266463752,0.9502905982905983,0.947008547008547,-0.0032820512820512793,0.9456351636388378,0.9450376991573128,-0.0005974644815249786
+5gnidd,0.25,0,6,TCPConnectScan,4010,0.602416918429003,0.6033282904689864,0.0009113720399833714,0.9945137157107232,0.9945137157107232,0.0,0.7503292568203198,0.7510357815442561,0.000706524723936286
+5gnidd,0.25,0,7,UDPFlood,91468,0.6164110305415532,0.6164068765204967,-4.154021056468338e-06,1.0,1.0,0.0,0.7626909479020746,0.7626877681285099,-3.17977356467658e-06
+5gnidd,0.25,0,8,UDPScan,3181,0.9860248447204969,0.9857187208941323,-0.00030612382636463487,0.9981138006916064,0.9981138006916064,0.0,0.9920324949226683,0.9918775382692908,-0.00015495665337750264
+5gnidd,0.25,1,0,Benign,95547,0.9997317596566524,0.9997317308724112,-2.8784241212775896e-08,0.39006980857588414,0.39002794436246035,-4.1864213423781216e-05,0.5611810851709367,0.5611377547563299,-4.3330414606801426e-05
+5gnidd,0.25,1,1,HTTPFlood,28162,0.8752899868330303,0.8513649228521866,-0.023925063980843708,0.9914068603082168,0.9933598466018039,0.001952986293587089,0.9297369297369297,0.9168974615296372,-0.012839468207292426
+5gnidd,0.25,1,2,ICMPFlood,231,0.3100671140939597,0.319060773480663,0.008993659386703268,1.0,1.0,0.0,0.47336065573770486,0.4837696335078534,0.010408977770148542
+5gnidd,0.25,1,3,SYNFlood,1944,0.900871459694989,0.8882921589688507,-0.012579300726138332,0.8508230452674898,0.8508230452674898,0.0,0.8751322751322751,0.8691539674198635,-0.00597830771241159
+5gnidd,0.25,1,4,SYNScan,4009,0.9844827586206897,0.9876451692611812,0.0031624106404914976,0.9970067348465952,0.9970067348465952,0.0,0.9907051679266329,0.9923038728897716,0.0015987049631387418
+5gnidd,0.25,1,5,SlowrateDoS,14625,0.9779650032404407,0.9814024390243903,0.003437435783949616,0.7222564102564103,0.6603076923076923,-0.061948717948718035,0.8308817745614725,0.7894543225015328,-0.041427452059939696
+5gnidd,0.25,1,6,TCPConnectScan,4010,0.7517904259329061,0.7534945221004912,0.0017040961675850452,0.9947630922693267,0.9947630922693267,0.0,0.8563761270931731,0.8574806534823731,0.0011045263892000001
+5gnidd,0.25,1,7,UDPFlood,91468,0.6178308239950826,0.6177640598934238,-6.676410165884494e-05,1.0,1.0,0.0,0.7637767989478739,0.7637257807966401,-5.101815123376152e-05
+5gnidd,0.25,1,8,UDPScan,3181,0.9906367041198502,0.9881693648816936,-0.002467339238156563,0.9977994341402074,0.9977994341402074,0.0,0.9942051683633515,0.9929610511496949,-0.0012441172136565815
+5gnidd,0.25,2,0,Benign,95547,0.9997829507840903,0.9997825496058712,-4.011782190893598e-07,0.38567406616638933,0.3849623745381854,-0.0007116916282039476,0.5566255050791133,0.5558838420094153,-0.0007416630696980286
+5gnidd,0.25,2,1,HTTPFlood,28162,0.9648802870757104,0.9619441923774955,-0.002936094698214875,0.5824160215893758,0.602265464100561,0.019849442511185256,0.726378955293284,0.7407520635891166,0.014373108295832604
+5gnidd,0.25,2,2,ICMPFlood,231,0.9058823529411765,0.90234375,-0.003538602941176472,1.0,1.0,0.0,0.9506172839506173,0.9486652977412731,-0.0019519862093442253
+5gnidd,0.25,2,3,SYNFlood,1944,0.8328387734915925,0.8390632785251619,0.006224505033569416,0.8662551440329218,0.8662551440329218,0.0,0.8492183560262229,0.8524424196405973,0.0032240636143744394
+5gnidd,0.25,2,4,SYNScan,4009,0.983779798476284,0.983779798476284,0.0,0.9985033674232976,0.9985033674232976,0.0,0.9910869026986877,0.9910869026986877,0.0
+5gnidd,0.25,2,5,SlowrateDoS,14625,0.5431091424796984,0.5538694388019277,0.010760296322229301,0.9557606837606838,0.9508376068376069,-0.004923076923076919,0.6926316832664385,0.6999899325480721,0.007358249281633533
+5gnidd,0.25,2,6,TCPConnectScan,4010,0.6829728801922417,0.6828556718723185,-0.0001172083199231233,0.9922693266832918,0.9922693266832918,0.0,0.8090687271248475,0.8089864796177696,-8.224750707785944e-05
+5gnidd,0.25,2,7,UDPFlood,91468,0.6171097017946297,0.6167809627846446,-0.00032873900998509065,1.0,1.0,0.0,0.7632255265178065,0.762974053977403,-0.00025147254040347455
+5gnidd,0.25,2,8,UDPScan,3181,0.9927922281416484,0.993103448275862,0.00031122013421369665,0.9959132348318139,0.9959132348318139,0.0,0.9943502824858758,0.9945063569298382,0.00015607444396248393
+5gnidd,0.5,0,0,Benign,95547,0.9999448838426985,0.9999183117767189,-2.6572065979668658e-05,0.3797607460202832,0.384334411336829,0.004573665316545794,0.5504649786853164,0.5552497883149873,0.004784809629670872
+5gnidd,0.5,0,1,HTTPFlood,28162,0.9479713051527306,0.8943990850462242,-0.05357222010650642,0.8727718201832256,0.9996804204246857,0.1269086002414601,0.9088186356073212,0.9441137510689314,0.03529511546161013
+5gnidd,0.5,0,2,ICMPFlood,231,0.75,0.8555555555555555,0.10555555555555551,1.0,1.0,0.0,0.8571428571428571,0.9221556886227544,0.06501283147989734
+5gnidd,0.5,0,3,SYNFlood,1944,0.8426909458775923,0.718103448275862,-0.12458749760173027,0.8569958847736625,0.8569958847736625,0.0,0.8497832185666921,0.7814258911819886,-0.06835732738470357
+5gnidd,0.5,0,4,SYNScan,4009,0.9908142999006951,0.997003745318352,0.006189445417656869,0.9955101022698928,0.9960089797954602,0.0004988775255674316,0.9931566504914769,0.9965061142999749,0.0033494638084979877
+5gnidd,0.5,0,5,SlowrateDoS,14625,0.7866928758637122,0.9995554765291608,0.2128626006654486,0.903042735042735,0.7687521367521367,-0.1342905982905983,0.8408620634769044,0.8690913307308777,0.028229267253973345
+5gnidd,0.5,0,6,TCPConnectScan,4010,0.6363201530612245,0.7178841309823678,0.08156397792114334,0.9952618453865336,0.9950124688279302,-0.00024937655860346464,0.7763081112624004,0.8340301003344481,0.05772198907204773
+5gnidd,0.5,0,7,UDPFlood,91468,0.6164733475767155,0.6166395879541302,0.00016624037741475028,1.0,1.0,0.0,0.7627386476874263,0.7628658762890896,0.00012722860166336503
+5gnidd,0.5,0,8,UDPScan,3181,0.9872512437810945,0.9796359148410985,-0.007615328939996058,0.9981138006916064,0.9981138006916064,0.0,0.9926528060028138,0.9887885393958268,-0.0038642666069870213
+5gnidd,0.5,1,0,Benign,95547,0.9999462495632777,0.999839232603628,-0.00010701695964976388,0.3894104472144599,0.3905407809769014,0.0011303337624415377,0.5605321040103649,0.5616852816328989,0.001153177622533974
+5gnidd,0.5,1,1,HTTPFlood,28162,0.9617002756419556,0.9897763578274761,0.02807608218552049,0.9415524465591932,0.8800511327320503,-0.06150131382714297,0.9515197186636524,0.9316942972068719,-0.019825421456780457
+5gnidd,0.5,1,2,ICMPFlood,231,0.9314516129032258,0.3285917496443812,-0.6028598632588446,1.0,1.0,0.0,0.964509394572025,0.49464668094218417,-0.46986271362984083
+5gnidd,0.5,1,3,SYNFlood,1944,0.8499486125385406,0.8646105593308939,0.014661946792353331,0.8508230452674898,0.8508230452674898,0.0,0.8503856041131106,0.8576613948664766,0.00727579075336604
+5gnidd,0.5,1,4,SYNScan,4009,0.9955223880597015,0.9901039089559623,-0.00541847910373916,0.9982539286605139,0.9982539286605139,0.0,0.9968862872088678,0.9941622158738045,-0.0027240713350633605
+5gnidd,0.5,1,5,SlowrateDoS,14625,0.8914421553090333,0.8090554519247075,-0.08238670338432574,0.9230769230769231,0.9786666666666667,0.05558974358974356,0.9069837750680237,0.8858150761232827,-0.02116869894474105
+5gnidd,0.5,1,6,TCPConnectScan,4010,0.6801909307875895,0.7623781303766011,0.08218719958901155,0.9950124688279302,0.9945137157107232,-0.0004987531172069293,0.8080194410692589,0.8631100530245644,0.05509061195530551
+5gnidd,0.5,1,7,UDPFlood,91468,0.6184114449522676,0.6179936219663802,-0.0004178229858874394,1.0,1.0,0.0,0.7642203061292695,0.7639011842522841,-0.000319121876985351
+5gnidd,0.5,1,8,UDPScan,3181,0.9733455882352942,0.9835811648079306,0.010235576572636407,0.9987425337944043,0.9981138006916064,-0.0006287331027978738,0.9858805275407293,0.9907941956623498,0.004913668121620529
+5gnidd,0.5,2,0,Benign,95547,0.9998089154587394,1.0,0.00019108454126059993,0.38332967021465875,0.38307848493411617,-0.00025118528054257627,0.5541836889090633,0.5539504650054106,-0.00023322390365276835
+5gnidd,0.5,2,1,HTTPFlood,28162,0.9545717888423133,0.9940183806052924,0.039446591762979044,0.9946026560613592,0.8910233648178396,-0.10357929124351961,0.9741761585949048,0.9397071490094746,-0.0344690095854302
+5gnidd,0.5,2,2,ICMPFlood,231,0.3869346733668342,0.8783269961977186,0.4913923228308844,1.0,1.0,0.0,0.5579710144927537,0.9352226720647774,0.3772516575720237
+5gnidd,0.5,2,3,SYNFlood,1944,0.7619909502262443,0.6519550909794812,-0.11003585924676307,0.8662551440329218,0.8662551440329218,0.0,0.810784785748676,0.7439805610779765,-0.06680422467069946
+5gnidd,0.5,2,4,SYNScan,4009,0.9849790691947796,0.9825724104074619,-0.002406658787317717,0.9977550511349463,0.9985033674232976,0.0007483162883512584,0.9913258983890954,0.990473833972535,-0.000852064416560383
+5gnidd,0.5,2,5,SlowrateDoS,14625,0.9887288198850489,0.8244942279117613,-0.1642345919732876,0.9057094017094017,0.9864615384615385,0.08075213675213677,0.9454000428234959,0.8982349095663543,-0.047165133257141556
+5gnidd,0.5,2,6,TCPConnectScan,4010,0.7177637511271415,0.7022927689594356,-0.015470982167705927,0.9925187032418953,0.9930174563591022,0.0004987531172069293,0.8330716902145473,0.8227272727272729,-0.010344417487274438
+5gnidd,0.5,2,7,UDPFlood,91468,0.6172304660876842,0.6177557153952656,0.0005252493075814657,1.0,1.0,0.0,0.7633178808223351,0.7637194040069636,0.0004015231846284939
+5gnidd,0.5,2,8,UDPScan,3181,0.990625,0.9968533668974198,0.00622836689741979,0.9965419679346118,0.9959132348318139,-0.0006287331027978738,0.9935746748158596,0.9963830791004876,0.00280840428462803
+5gnidd,0.75,0,0,Benign,95547,0.9998897068960764,0.9998905758446177,8.689485413437836e-07,0.3795304928464525,0.38254471621296327,0.0030142233665107487,0.5502146964662327,0.5533754220223767,0.003160725556143995
+5gnidd,0.75,0,1,HTTPFlood,28162,0.9621111033078166,0.955288165720927,-0.006822937586889566,0.9729067537816917,0.9923300901924579,0.019423336410766212,0.9674788135593221,0.9734568761320884,0.0059780625727663406
+5gnidd,0.75,0,2,ICMPFlood,231,0.7993079584775087,0.7356687898089171,-0.06363916866859154,1.0,1.0,0.0,0.8884615384615385,0.8477064220183486,-0.040755116443189965
+5gnidd,0.75,0,3,SYNFlood,1944,0.8474059003051883,0.8138739618954568,-0.033531938409731454,0.8569958847736625,0.8569958847736625,0.0,0.8521739130434783,0.8348784765722876,-0.017295436471190695
+5gnidd,0.75,0,4,SYNScan,4009,0.9883749690823646,0.9937826411340462,0.005407672051681689,0.9967572960838115,0.9967572960838115,0.0,0.992548435171386,0.9952677459526775,0.0027193107812915063
+5gnidd,0.75,0,5,SlowrateDoS,14625,0.9466956949223962,0.9842414331375902,0.03754573821519391,0.9217094017094017,0.9053675213675214,-0.016341880341880333,0.9340354767184035,0.9431583446114397,0.00912286789303618
+5gnidd,0.75,0,6,TCPConnectScan,4010,0.6302733449202086,0.6894227445558244,0.05914939963561583,0.9947630922693267,0.9947630922693267,0.0,0.771641357964987,0.814414046549612,0.04277268858462502
+5gnidd,0.75,0,7,UDPFlood,91468,0.6164982779189443,0.6158548901846191,-0.0006433877343251515,1.0,1.0,0.0,0.762757729272208,0.7622650943789325,-0.0004926348932754454
+5gnidd,0.75,0,8,UDPScan,3181,0.9918724601437949,0.9844961240310077,-0.007376336112787207,0.9974850675888085,0.9981138006916064,0.0006287331027978738,0.9946708463949843,0.9912581954417733,-0.0034126509532109717
+5gnidd,0.75,1,0,Benign,95547,0.9997597629661266,0.9998132685997813,5.350563365469352e-05,0.3919955623933771,0.39226767978063154,0.0002721173872544669,0.5631757010751072,0.5634649788775802,0.00028927780247300916
+5gnidd,0.75,1,1,HTTPFlood,28162,0.9582778062532035,0.9600082174895569,0.001730411236353424,0.9958099566792131,0.9955969036290037,-0.0002130530502094663,0.97668344158671,0.9774787337888718,0.0007952922021617903
+5gnidd,0.75,1,2,ICMPFlood,231,0.3442622950819672,0.673469387755102,0.3292070926731348,1.0,1.0,0.0,0.5121951219512195,0.8048780487804877,0.2926829268292682
+5gnidd,0.75,1,3,SYNFlood,1944,0.864158829676071,0.8512609366958311,-0.012897892980239911,0.8508230452674898,0.8508230452674898,0.0,0.8574390876101607,0.8510419346539748,-0.006397152956185925
+5gnidd,0.75,1,4,SYNScan,4009,0.9918114143920596,0.9952724558347847,0.003461041442725188,0.9970067348465952,0.9977550511349463,0.0007483162883511474,0.9944022888418957,0.996512207274539,0.0021099184326432985
+5gnidd,0.75,1,5,SlowrateDoS,14625,0.991245641368054,0.9907674126597238,-0.00047822870833025277,0.9135726495726496,0.9171965811965812,0.0036239316239315977,0.9508255052661543,0.9525635563130238,0.0017380510468695265
+5gnidd,0.75,1,6,TCPConnectScan,4010,0.757690846942651,0.7281854691493246,-0.029505377793326404,0.9950124688279302,0.9947630922693267,-0.00024937655860346464,0.8602846054333765,0.8408516020236088,-0.019433003409767724
+5gnidd,0.75,1,7,UDPFlood,91468,0.6186373043678223,0.6183570960174688,-0.00028020835035347247,1.0,1.0,0.0,0.7643927428318333,0.7641788052082594,-0.0002139376235739121
+5gnidd,0.75,1,8,UDPScan,3181,0.9811437403400309,0.9793273680962666,-0.001816372243764297,0.9977994341402074,0.9977994341402074,0.0,0.9894014962593516,0.9884771099345998,-0.0009243863247517803
+5gnidd,0.75,2,0,Benign,95547,1.0,1.0,0.0,0.3882382492385946,0.38955697196144307,0.0013187227228484422,0.559325100646854,0.5606923355025307,0.0013672348556766911
+5gnidd,0.75,2,1,HTTPFlood,28162,0.9646323580755534,0.9567800081933634,-0.007852349882189968,0.9810737873730558,0.9951707975285846,0.014097010155528777,0.9727836067882544,0.9755978696000279,0.002814262811773549
+5gnidd,0.75,2,2,ICMPFlood,231,0.4103019538188277,0.4502923976608187,0.039990443841990964,1.0,1.0,0.0,0.5818639798488665,0.6209677419354839,0.039103762086617366
+5gnidd,0.75,2,3,SYNFlood,1944,0.883989501312336,0.8779979144942649,-0.005991586818071104,0.8662551440329218,0.8662551440329218,0.0,0.8750324759677839,0.8720870015535992,-0.00294547441418469
+5gnidd,0.75,2,4,SYNScan,4009,0.9977539306214125,0.9972575417601596,-0.0004963888612529166,0.9972561736093789,0.9977550511349463,0.0004988775255674316,0.9975049900199601,0.9975062344139651,1.244394005062155e-06
+5gnidd,0.75,2,5,SlowrateDoS,14625,0.9616695059625213,0.9895190663792462,0.027849560416724928,0.9263589743589744,0.9102222222222223,-0.01613675213675214,0.9436840455542786,0.9482156848778402,0.004531639323561598
+5gnidd,0.75,2,6,TCPConnectScan,4010,0.7491538172245205,0.7497177267595032,0.0005639095349827539,0.9935162094763093,0.9935162094763093,0.0,0.8542024013722126,0.8545688545688546,0.0003664531966419693
+5gnidd,0.75,2,7,UDPFlood,91468,0.6166853197772414,0.6171430113621029,0.0004576915848615304,1.0,1.0,0.0,0.7629008715959797,0.7632510013351135,0.000350129739133731
+5gnidd,0.75,2,8,UDPScan,3181,0.9793018226753166,0.9774899784150478,-0.0018118442602688134,0.9965419679346118,0.9965419679346118,0.0,0.9878466812090994,0.9869240348692404,-0.0009226463398589457
@@ -0,0 +1,181 @@
+dataset,alpha,seed,selected_beta,client_id,n_samples,orig_rare_f1,prop_rare_f1,delta_rare_f1,orig_rare_recall,prop_rare_recall,delta_rare_recall,orig_macro_f1,prop_macro_f1,delta_macro_f1,orig_accuracy,prop_accuracy,delta_accuracy,status
+xiiotid,0.25,0,0.0,0,10103,0.46517066665428247,0.4671001573063932,0.0019294906521107325,0.8052377497488559,0.8052918154928006,5.406574394473118e-05,0.5151407310607025,0.5161551340849532,0.0010144030242507096,0.6685143026823716,0.6686132831832129,9.898050084133914e-05,unchanged
+xiiotid,0.25,0,0.0,1,33050,0.8526858188186391,0.85459860592547,0.0019127871068308488,0.7732930685256398,0.775599493566826,0.0023064250411861664,0.6094600336361348,0.6102150195625664,0.0007549859264316394,0.941391830559758,0.9421785173978819,0.0007866868381239955,unchanged
+xiiotid,0.25,0,0.0,2,3306,0.01785714285714286,0.01785714285714286,0.0,1.0,1.0,0.0,0.6351628803498259,0.6354582484509493,0.0002953681011234055,0.6536600120992135,0.6536600120992135,0.0,unchanged
+xiiotid,0.25,0,0.0,3,4060,0.30471864281079075,0.3296643810898593,0.02494573827906854,0.4696969696969697,0.5,0.030303030303030276,0.7511565095776971,0.7583823555044171,0.0072258459267200426,0.8847290640394089,0.884975369458128,0.00024630541871917266,improved
+xiiotid,0.25,0,0.0,4,84496,0.04477127207787965,0.04514717952800433,0.0003759074501246823,0.7765086206896552,0.7799568965517242,0.0034482758620689724,0.6091043195772301,0.609638466061856,0.0005341464846259081,0.8875923120621094,0.8885864419617496,0.000994129899640206,unchanged
+xiiotid,0.25,0,0.0,5,11466,0.6707020276472061,0.6726155893322339,0.0019135616850277648,0.9733183295823955,0.9733183295823955,0.0,0.8093164577147938,0.812282294679531,0.0029658369647371785,0.840572126286412,0.84135705564277,0.0007849293563579884,unchanged
+xiiotid,0.25,0,0.0,6,1423,0.8799999999999999,0.8799999999999999,0.0,0.9166666666666666,0.9166666666666666,0.0,0.6880530092947279,0.6880530092947279,0.0,0.7940969782150387,0.7940969782150387,0.0,unchanged
+xiiotid,0.25,0,0.0,7,7122,0.803103263777421,0.8073593073593074,0.0042560435818863596,0.8604368932038835,0.8604368932038835,0.0,0.6239407333483753,0.6252586433720333,0.0013179100236579977,0.9727604605447908,0.9731816905363662,0.0004212299915753448,unchanged
+xiiotid,0.25,0,0.0,8,6566,0.4279675955930006,0.4312875704523096,0.0033199748593089606,0.7452053626876758,0.7474519927425933,0.002246630054917542,0.5698064534966384,0.5712841418853772,0.0014776883887388514,0.7281449893390192,0.7298202863234846,0.0016752969844654464,unchanged
+xiiotid,0.25,0,0.0,9,1239,0.9758454106280192,0.9758454106280192,0.0,0.9805825242718447,0.9805825242718447,0.0,0.7130398411872055,0.7130398411872055,0.0,0.9475383373688459,0.9475383373688459,0.0,unchanged
+xiiotid,0.25,1,0.0,0,4636,0.7766093179021245,0.7715460489192179,-0.005063268982906655,0.7622910521140609,0.7593411996066863,-0.002949852507374562,0.7161778391450923,0.714099813943949,-0.0020780252011433076,0.7340379637618637,0.7312338222605694,-0.002804141501294266,unchanged
+xiiotid,0.25,1,0.0,1,38257,0.0027173913043478256,0.002680965147453083,-3.642615689474242e-05,1.0,1.0,0.0,0.6702493195383108,0.6705126557034797,0.000263336165168937,0.8903991426405625,0.8844655879969678,-0.005933554643594685,unchanged
+xiiotid,0.25,1,0.0,2,6385,0.6730025773195876,0.6730025773195876,0.0,0.8220338983050848,0.8220338983050848,0.0,0.7640030931538621,0.764104048652251,0.0001009554983889549,0.9617854346123728,0.9617854346123728,0.0,unchanged
+xiiotid,0.25,1,0.0,3,22620,0.7121212121212122,0.6937269372693726,-0.01839427485183953,0.9894736842105263,0.9894736842105263,0.0,0.7268252099092205,0.7243730725436959,-0.0024521373655246315,0.8232537577365163,0.8231211317418214,-0.00013262599469487935,unchanged
+xiiotid,0.25,1,0.0,4,16888,0.1367694805194805,0.13014311270125226,-0.006626367818228257,0.6741854636591479,0.6741854636591479,0.0,0.663814741271609,0.6602136391863644,-0.0036011020852445563,0.9104097584083373,0.9062055897678826,-0.004204168640454697,unchanged
+xiiotid,0.25,1,0.0,5,38817,0.9993425378040763,0.9993425378040763,0.0,0.9986859395532195,0.9986859395532195,0.0,0.695932201971511,0.6932726391940013,-0.002659562777509672,0.8820362212432697,0.8756730298580518,-0.006363191385217903,unchanged
+xiiotid,0.25,1,0.0,6,13767,0.5453152519452761,0.5381491560424561,-0.007166095902820024,0.8130193905817175,0.8130193905817175,0.0,0.5726111792468381,0.569979314683699,-0.00263186456313913,0.8192053461175274,0.8168809471925619,-0.0023243989249654984,unchanged
+xiiotid,0.25,1,0.0,7,2735,0.9380619380619379,0.9380619380619379,0.0,0.9838709677419355,0.9838709677419355,0.0,0.7769170844935405,0.7761324111935268,-0.0007846733000136386,0.9224862888482632,0.9221206581352833,-0.000365630712979903,unchanged
+xiiotid,0.25,1,0.0,8,10542,0.48719975324825543,0.4871977240398293,-2.029208426124196e-06,1.0,1.0,0.0,0.737824227086807,0.7373451704100447,-0.0004790566767622906,0.7452096376399165,0.7422690191614495,-0.0029406184784670497,unchanged
+xiiotid,0.25,1,0.0,9,8184,0.1237785016286645,0.1237785016286645,0.0,0.6129032258064516,0.6129032258064516,0.0,0.706637162220372,0.7050119981335593,-0.0016251640868126804,0.9445259042033236,0.9442815249266863,-0.00024437927663734094,unchanged
+xiiotid,0.25,2,0.0,0,36328,0.8731533367294957,0.870096790626592,-0.0030565461029037344,0.9861910241657077,0.9827387802071347,-0.0034522439585730202,0.7316551446485546,0.7309928867235244,-0.000662257925030274,0.8910757542391544,0.8902224179696102,-0.0008533362695442026,unchanged
+xiiotid,0.25,2,0.0,1,21015,0.6827751250183436,0.679480525278908,-0.0032945997394355775,0.8593951827368741,0.8593951827368741,0.0,0.6612426980344254,0.6595639153476706,-0.0016787826867548095,0.9340471092077087,0.9311444206519153,-0.0029026885557934534,unchanged
+xiiotid,0.25,2,0.0,2,1307,0.0,0.0,0.0,0.0,0.0,0.0,0.8491288720554245,0.8359794844458573,-0.013149387609567187,0.9556235654169855,0.9556235654169855,0.0,unchanged
+xiiotid,0.25,2,0.0,3,13343,0.8108463521119351,0.7845690894471382,-0.026277262664796908,0.8929944607363962,0.8905506679700228,-0.0024437927663734094,0.5855922277675589,0.5770334707682923,-0.00855875699926667,0.722925878737915,0.7252491943341077,0.002323315596192699,degraded
+xiiotid,0.25,2,0.0,4,5069,0.8348416289592759,0.8331441543700341,-0.0016974745892418541,0.727810650887574,0.7238658777120316,-0.0039447731755424265,0.82046158924311,0.8233029797675057,0.0028413905243956927,0.8046952061550602,0.806865259420004,0.0021700532649437942,unchanged
+xiiotid,0.25,2,0.0,5,6963,0.8747839356315461,0.8720184662734034,-0.0027654693581427026,0.7813606429684181,0.7772801302931596,-0.004080512675258552,0.7262748617372359,0.7257893380347387,-0.00048552370249721477,0.9302024989228781,0.9279046388051128,-0.002297860117765338,unchanged
+xiiotid,0.25,2,0.0,6,14944,0.44319659072026896,0.44167452459531664,-0.0015220661249523149,0.751531862745098,0.751531862745098,0.0,0.5574556521660132,0.553401720969503,-0.004053931196510208,0.9022350107066381,0.9000267665952891,-0.0022082441113490114,unchanged
+xiiotid,0.25,2,0.0,7,16849,0.9168539325842696,0.9168539325842696,0.0,1.0,1.0,0.0,0.6709137398535587,0.6664470168261388,-0.0044667230274199055,0.9196391477239005,0.9170870674817496,-0.0025520802421509226,unchanged
+xiiotid,0.25,2,0.0,8,38722,0.9132624800941633,0.9026922890515194,-0.010570191042643895,0.8810171568627452,0.8673406862745098,-0.013676470588235401,0.6642749317775651,0.6622265186968516,-0.0020484130807135736,0.9249780486545116,0.9224471876452662,-0.0025308610092453776,unchanged
+xiiotid,0.25,2,0.0,9,8291,0.7901491414964281,0.7924335122986639,0.002284370802235758,0.7317030409286016,0.7347862886161658,0.0030832476875641834,0.7490718105429928,0.7486309482509236,-0.0004408622920691885,0.7112531660837053,0.7160776745869014,0.0048245085031961565,unchanged
+xiiotid,0.5,0,0.75,0,22614,0.4990313061051025,0.49852708011546776,-0.0005042259896347323,0.9921372408863474,0.9305694543721705,-0.061567786514176825,0.606034259972128,0.6157366471089164,0.009702387136788393,0.8891394711240824,0.8903334217741222,0.001193950650039799,unchanged
+xiiotid,0.5,0,0.75,1,33759,0.38653536479623435,0.3950598098818963,0.008524445085661969,0.6685349322210636,0.6940823774765381,0.025547445255474477,0.6496912815143935,0.6616396943744374,0.011948412860043889,0.9208803578305045,0.9189253236174058,-0.0019550342130987275,unchanged
+xiiotid,0.5,0,0.75,2,5625,0.876107292183403,0.8949207257273619,0.018813433543958924,1.0,0.952,-0.04800000000000004,0.7720803645045097,0.7766807853440173,0.004600420839507624,0.9196444444444445,0.8968888888888888,-0.02275555555555564,unchanged
+xiiotid,0.5,0,0.75,3,13539,0.6162162162162163,0.35962145110410093,-0.2565947651121153,0.5757575757575758,0.5757575757575758,0.0,0.8402794288226,0.7905690740085016,-0.049710354814098356,0.9861880493389468,0.9793190043577812,-0.006869044981165584,degraded
+xiiotid,0.5,0,0.75,4,11881,0.1352659822678488,0.15006641568143958,0.014800433413590774,0.7972972972972974,0.7972972972972974,0.0,0.5992919833207005,0.6440684296718677,0.044776446351167176,0.6907667704738658,0.7001094183991247,0.009342647925258851,unchanged
+xiiotid,0.5,0,0.75,5,12785,0.6976090014064699,0.8338983050847457,0.13628930367827585,0.9959839357429718,0.9879518072289156,-0.008032128514056214,0.7629091141568864,0.7926623902258025,0.02975327606891609,0.9062964411419633,0.9104419241298397,0.0041454829878764254,improved
+xiiotid,0.5,0,0.75,6,6131,0.36292270531400966,0.4477565595404441,0.08483385422643441,1.0,1.0,0.0,0.7052693971614308,0.7195049733487044,0.014235576187273669,0.931821888762029,0.9106181699559615,-0.02120371880606753,improved
+xiiotid,0.5,0,0.75,7,34434,0.04508409872548376,0.03969509578636252,-0.005389002939121236,0.8492063492063492,0.8174603174603174,-0.031746031746031744,0.6087392282025439,0.6169939163182777,0.008254688115733821,0.9070685949933206,0.904571063483766,-0.0024975315095545625,unchanged
+xiiotid,0.5,0,0.75,8,5090,0.6294201152886518,0.6208429867121025,-0.008577128576549353,0.975206611570248,0.9487898465171193,-0.026416765053128732,0.8028671094844209,0.7924831635124538,-0.010383945971967012,0.9589390962671905,0.9469548133595285,-0.01198428290766207,unchanged
+xiiotid,0.5,0,0.75,9,16973,0.38940807646284775,0.400258937520882,0.010850861058034234,0.6283261802575106,0.6476394849785407,0.0193133047210301,0.5518204012277159,0.5718262874717445,0.020005886244028548,0.8041006304130089,0.8091675013256349,0.00506687091262592,unchanged
+xiiotid,0.5,1,0.75,0,7041,0.1371841155234657,0.14925373134328357,0.012069615819817858,0.95,0.75,-0.19999999999999996,0.679239688702505,0.6893802671721605,0.010140578469655548,0.8568385172560716,0.8953273682715523,0.03848885101548072,unchanged
+xiiotid,0.5,1,0.75,1,8656,0.5400867572182313,0.5384279141942322,-0.0016588430239991014,1.0,0.8333333333333333,-0.16666666666666674,0.6928923604930376,0.7095521010249697,0.016659740531932088,0.8959103512014788,0.9071164510166358,0.011206099815157078,unchanged
+xiiotid,0.5,1,0.75,2,65841,0.3334102074903383,0.374319944106564,0.04090973661622571,0.8706597222222222,0.8625578703703703,-0.00810185185185186,0.6856401858050954,0.7114075329715898,0.025767347166494314,0.9130784769368631,0.9370453061162497,0.02396682917938664,improved
+xiiotid,0.5,1,0.75,3,12392,0.4784482758620689,0.4691176470588235,-0.009330628803245422,0.75,0.7222222222222222,-0.02777777777777779,0.7593818733983905,0.7501539859855098,-0.00922788741288072,0.9825693996126533,0.9825693996126533,0.0,unchanged
+xiiotid,0.5,1,0.75,4,9970,0.48653657481685963,0.5099275046378224,0.023390929820962802,0.6756830601092896,0.6649882903981265,-0.010694769711163121,0.5749858563972289,0.6030572290400724,0.02807137264284354,0.8415245737211635,0.8841524573721163,0.042627883650952825,improved
+xiiotid,0.5,1,0.75,5,17906,0.5121799124939562,0.5207373271889401,0.008557414694983834,0.8828929171807846,0.8823008190001408,-0.0005920981806437631,0.6956452071069839,0.7050729428085704,0.009427735701586504,0.9187981682117726,0.9391823969619122,0.020384228750139566,unchanged
+xiiotid,0.5,1,0.75,6,5665,0.6686423544465772,0.6665841046919885,-0.0020582497545886813,0.8171191553544495,0.816157616892911,-0.0009615384615384581,0.7508169025076517,0.7612280152503897,0.010411112742738093,0.946513680494263,0.9548102383053839,0.00829655781112093,unchanged
+xiiotid,0.5,1,0.75,7,20619,0.495821276210082,0.495522836079836,-0.0002984401302459916,0.6952614379084967,0.6919934640522876,-0.0032679738562091387,0.6625506568546684,0.6819169023448449,0.019366245490176515,0.817886415442068,0.8648334060817693,0.046946990639701314,unchanged
+xiiotid,0.5,1,0.75,8,10808,0.18449197860962568,0.2456165984804208,0.06112461987079512,0.5535714285714286,0.625,0.0714285714285714,0.5955973042339031,0.6140823923870982,0.01848508815319505,0.9074759437453738,0.929219096965211,0.02174315321983722,improved
+xiiotid,0.5,1,0.75,9,3933,0.9471753788453319,0.9596790172379631,0.012503638392631244,0.9797101449275363,0.980952380952381,0.001242236024844745,0.8165318686699652,0.8258516490377947,0.009319780367829544,0.9494024917365879,0.954741927281973,0.005339435545385185,unchanged
+xiiotid,0.5,2,0.75,0,8490,0.29545454545454547,0.6190476190476191,0.3235930735930736,0.6190476190476191,0.6190476190476191,0.0,0.8108503184502494,0.8379133763903672,0.027063057940117874,0.9479387514723204,0.9445229681978798,-0.003415783274440587,improved
+xiiotid,0.5,2,0.75,1,22974,0.44292423142096343,0.47183404955170033,0.028909818130736897,0.6950554798520537,0.7582246447342806,0.0631691648822269,0.7296833677456545,0.7213598946985074,-0.00832347304714709,0.9630887089753635,0.93187951597458,-0.03120919300078351,improved
+xiiotid,0.5,2,0.75,2,7058,0.40458786474118974,0.4850632911392405,0.08047542639805078,0.7601626016260162,0.7235772357723578,-0.03658536585365846,0.7586134850061949,0.7652645783883212,0.006651093382126327,0.9474355341456503,0.9397846415415132,-0.007650892604137094,improved
+xiiotid,0.5,2,0.75,3,13110,0.8069105691056911,0.90177736202058,0.09486679291488886,0.7026548672566372,0.8530973451327434,0.1504424778761062,0.6828979964087885,0.7603952204580291,0.07749722404924064,0.7835240274599542,0.8723112128146453,0.0887871853546911,improved
+xiiotid,0.5,2,0.75,4,26172,0.47994748655401687,0.48792156862745095,0.007974082073434086,0.753968253968254,0.753968253968254,0.0,0.6480479243341395,0.6878510221739338,0.0398030978397943,0.8920984257985634,0.8923276784349686,0.00022925263640527493,unchanged
+xiiotid,0.5,2,0.75,5,18577,0.00554016620498615,0.03636363636363636,0.030823470158650213,1.0,1.0,0.0,0.7087202635739932,0.6986678864522665,-0.010052377121726686,0.9296441836679765,0.9036981213328309,-0.025946062335145625,improved
+xiiotid,0.5,2,0.75,6,10875,0.9786258988427716,0.9884467698544801,0.009820871011708476,0.9948990232902789,0.9800117856732207,-0.014887237617058235,0.7480731033355137,0.7725964432361911,0.024523339900677454,0.9119080459770115,0.9406896551724138,0.02878160919540229,unchanged
+xiiotid,0.5,2,0.75,7,17904,0.4126664831558219,0.39871273712737126,-0.01395374602845062,0.9142857142857144,0.8718253968253968,-0.04246031746031753,0.6049295014182259,0.6472344346590341,0.042304933240808196,0.9354892761394102,0.8895218945487042,-0.045967381590706013,unchanged
+xiiotid,0.5,2,0.75,8,12456,0.31673959266955587,0.46450943724005433,0.14776984457049847,0.8279714030384271,0.8199285075960678,-0.008042895442359255,0.6870896919114862,0.7233149508993455,0.03622525898785933,0.7365928066795119,0.855892742453436,0.11929993577392417,improved
+xiiotid,0.5,2,0.75,9,25215,0.2679012345679012,0.3168406402226861,0.048939405654784884,0.7598290598290598,0.6965811965811965,-0.06324786324786325,0.6555531744563652,0.6622507884429573,0.006697613986592077,0.9558992663097363,0.9085068411659727,-0.04739242514376363,improved
+xiiotid,0.75,0,0.5,0,13683,0.4177077040795779,0.42980808493262557,0.012100380853047654,0.8869463869463869,0.8927738927738929,0.005827505827505908,0.7937995365037241,0.7927057609904862,-0.0010937755132379712,0.8982679237009428,0.8996565080757144,0.0013885843747716287,unchanged
+xiiotid,0.75,0,0.5,1,25011,0.3121910695742472,0.34068209223488727,0.028491022660640075,0.6452067669172932,0.6715225563909775,0.026315789473684292,0.6687805910767683,0.6707862676702752,0.002005676593506811,0.9207148854504018,0.920794850265883,7.996481548122603e-05,improved
+xiiotid,0.75,0,0.5,2,6172,0.4697560975609756,0.49453759198846825,0.024781494427492667,0.6586206896551725,0.668144499178982,0.00952380952380949,0.8068507443087385,0.8009596997959896,-0.005891044512748911,0.931950745301361,0.9321127673363577,0.00016202203499671164,improved
+xiiotid,0.75,0,0.5,3,7526,0.43572770874821193,0.4459353986551393,0.010207689906927386,0.6972002434570906,0.6877662811929397,-0.009433962264150941,0.74878228028787,0.7514883414257376,0.0027060611378676436,0.907919213393569,0.9105766675524847,0.0026574541589157175,unchanged
+xiiotid,0.75,0,0.5,4,17179,0.5017536620590056,0.48478262983382436,-0.01697103222518126,0.8070175438596491,0.7701262502049516,-0.036891293654697455,0.586794036513965,0.5718644091286076,-0.014929627385357414,0.9561674137027767,0.9453984515978812,-0.01076896210489553,unchanged
+xiiotid,0.75,0,0.5,5,15317,0.8123249299719887,0.8454810495626821,0.03315611959069342,0.7754010695187166,0.7754010695187166,0.0,0.9149214374359469,0.9227559474175965,0.007834509981649584,0.9701638702095711,0.9697068616569824,-0.000457008552588678,improved
+xiiotid,0.75,0,0.5,6,12085,0.42435897435897435,0.4400552192842731,0.015696244925298752,0.8732394366197183,0.8873239436619718,0.014084507042253502,0.7006953696159909,0.707635506764101,0.006940137148110148,0.8714935870914357,0.8778651220521307,0.006371534960695091,unchanged
+xiiotid,0.75,0,0.5,7,10796,0.20254119285216987,0.2646949481597753,0.06215375530760542,0.5119047619047619,0.4841269841269842,-0.02777777777777768,0.6482071380843585,0.6603627848172349,0.01215564673287639,0.889866617265654,0.8894034827713968,-0.0004631344942571669,improved
+xiiotid,0.75,0,0.5,8,31158,0.006510764096243685,0.006082929643070903,-0.0004278344531727825,0.8125,0.8125,0.0,0.6718590534625448,0.6792150463009095,0.007355992838364633,0.9043584312215162,0.9070543680595674,0.0026959368380512494,unchanged
+xiiotid,0.75,0,0.5,9,23904,0.453612809784358,0.45119047619047614,-0.0024223335938818313,0.8913043478260869,0.8956521739130434,0.004347826086956497,0.6388011681801342,0.6331740214900163,-0.005627146690117879,0.8868390227576974,0.8941599732262383,0.0073209504685408655,unchanged
+xiiotid,0.75,1,0.5,0,27225,0.45870917042503756,0.4702134437719019,0.011504273346864324,0.7274545454545455,0.7194545454545455,-0.008000000000000007,0.7133569490706955,0.7291577221888712,0.015800773118175737,0.9441322314049587,0.9481726354453627,0.004040404040403955,unchanged
+xiiotid,0.75,1,0.5,1,3671,0.4597156398104265,0.45631067961165045,-0.0034049601987760636,0.4369369369369369,0.42342342342342343,-0.013513513513513487,0.7270015253911712,0.7372595364301502,0.010258011038978943,0.8430945246526832,0.8907654590029964,0.04767093435031322,unchanged
+xiiotid,0.75,1,0.5,2,14559,0.45427115350820474,0.4644647225470606,0.010193569038855876,0.7252226093689509,0.7089624467673248,-0.016260162601626105,0.7126051284771984,0.7263154752296265,0.013710346752428104,0.8658561714403462,0.8975204340957483,0.03166426265540212,unchanged
+xiiotid,0.75,1,0.5,3,14100,0.44281305526980264,0.451033302619455,0.008220247349652376,0.7363636363636363,0.7181818181818181,-0.018181818181818188,0.7088542984119168,0.7221862673213896,0.013331968909472747,0.9142553191489362,0.9386524822695036,0.02439716312056739,unchanged
+xiiotid,0.75,1,0.5,4,26763,0.5935552515579875,0.6074174966658153,0.013862245107827853,0.9685741088180113,0.9746716697936211,0.0060975609756097615,0.7359187499647767,0.7516990455997716,0.015780295634994945,0.9127526809401039,0.9337144565258005,0.020961775585696674,unchanged
+xiiotid,0.75,1,0.5,5,13029,0.49813859141302747,0.5007996698137543,0.0026610784007268795,0.8309748427672956,0.8309748427672956,0.0,0.6296355751387946,0.6435552476195708,0.01391967248077619,0.8740501957172462,0.9084350295494665,0.03438483383222035,unchanged
+xiiotid,0.75,1,0.5,6,28319,0.39883988993827624,0.42037598544572474,0.021536095507448505,0.7403474903474904,0.7403474903474904,0.0,0.722315800597633,0.7369713756969668,0.014655575099333795,0.9249973516013984,0.9399696316960344,0.014972280094636026,improved
+xiiotid,0.75,1,0.5,7,10991,0.06899625601711536,0.07465101108936725,0.005654755072251896,1.0,1.0,0.0,0.6351705801550876,0.6512421805693148,0.016071600414227172,0.9402238194886725,0.9449549631516696,0.004731143662997095,unchanged
+xiiotid,0.75,1,0.5,8,13070,0.4225040650406504,0.4393993458221826,0.016895280781532185,0.673890608875129,0.6475748194014448,-0.02631578947368418,0.7449117767273183,0.7606241166549877,0.015712339927669405,0.9169854628921194,0.9364192807957153,0.019433817903595996,unchanged
+xiiotid,0.75,1,0.5,9,11104,0.34860329372524496,0.4113137204792472,0.06271042675400224,0.8725490196078431,0.8686120117338274,-0.003937007874015741,0.7936261461178652,0.8138613481393286,0.02023520202146345,0.9172370317002881,0.9368695965417867,0.019632564841498557,improved
+xiiotid,0.75,2,0.5,0,10586,0.45646130909440674,0.47030303030303033,0.013841721208623592,0.7833333333333334,0.8125,0.029166666666666563,0.7387139319200966,0.7473809137907063,0.008666981870609658,0.9348195730209711,0.9388815416587947,0.004061968637823599,unchanged
+xiiotid,0.75,2,0.5,1,16425,0.026997578692493945,0.02867932231067565,0.0016817436181817032,0.6571428571428571,0.6571428571428571,0.0,0.6639395594587535,0.6588652395956793,-0.0050743198630741615,0.9580517503805175,0.9594520547945206,0.001400304414003073,unchanged
+xiiotid,0.75,2,0.5,2,7893,0.8550943396226416,0.8591819286324979,0.004087589009856285,0.9668641889242299,0.9379057557404906,-0.02895843318373925,0.780941698710889,0.7885002002775201,0.007558501566631071,0.9071329025718992,0.9110604332953249,0.003927530723425776,unchanged
+xiiotid,0.75,2,0.5,3,10387,0.15616194572148895,0.12441950464396284,-0.03174244107752611,0.8389830508474576,0.8389830508474576,0.0,0.7364533682211447,0.7307127161765188,-0.005740652044625927,0.8876480215654183,0.8899586020987773,0.0023105805333589746,degraded
+xiiotid,0.75,2,0.5,4,25057,0.044989734221148506,0.05532180332232777,0.010332069101179266,0.7,0.7,0.0,0.6953391950856829,0.6933656514701521,-0.001973543615530815,0.9151135411262322,0.9196631679770124,0.004549626850780197,unchanged
+xiiotid,0.75,2,0.5,5,8364,0.49603174603174605,0.42627906976744184,-0.0697526762643042,0.6783536585365854,0.6905487804878049,0.012195121951219523,0.7985669296874586,0.78131892014619,-0.017248009541268594,0.9577953132472501,0.953012912482066,-0.004782400765184103,degraded
+xiiotid,0.75,2,0.5,6,7987,0.2292975442604226,0.2350021815008726,0.005704637240450006,0.7317813765182186,0.7317813765182186,0.0,0.7159763619772302,0.7420719685266148,0.026095606549384587,0.9404031551270815,0.9451608864404658,0.004757731313384306,unchanged
+xiiotid,0.75,2,0.5,7,37651,0.9647773279352226,0.9560197910940077,-0.008757536841214941,0.9514944939695857,0.9349764027267959,-0.0165180912427898,0.7088047912490114,0.7103595834778935,0.001554792228882107,0.9251281506467292,0.9283153169902526,0.003187166343523451,unchanged
+xiiotid,0.75,2,0.5,8,26560,0.7272727272727273,0.7948717948717949,0.06759906759906764,0.7368421052631579,0.8157894736842105,0.07894736842105265,0.8243030196888212,0.834936911715183,0.01063389202636178,0.9631400602409639,0.965512048192771,0.002371987951807153,improved
+xiiotid,0.75,2,0.5,9,11921,0.5653650254668929,0.606023102310231,0.04065807684333811,0.8403508771929824,0.8605263157894737,0.02017543859649129,0.8542727132185379,0.8645597171004041,0.010287003881866208,0.9479909403573525,0.9510108212398288,0.003019880882476289,improved
+5gnidd,0.25,0,0.0,0,15707,0.9556097560975609,0.9551656920077973,-0.00044406408976360723,0.9978678038379531,0.9978678038379531,0.0,0.876005750846728,0.8711743108768858,-0.004831439969842255,0.8899216909658114,0.8895396956770867,-0.00038199528872473465,unchanged
+5gnidd,0.25,0,0.0,1,11805,0.8988283325772795,0.8935766190784538,-0.005251713498825694,0.9615995794353441,0.9599507748186913,-0.0016488046166528436,0.7808430139232388,0.7743678463814608,-0.006475167541778015,0.6813214739517154,0.6804743752647183,-0.0008470986869970609,unchanged
+5gnidd,0.25,0,0.0,2,62982,0.1163394220140022,0.10208948022787583,-0.014249941786126374,1.0,1.0,0.0,0.6565301672757924,0.6447101682710856,-0.01181999900470676,0.38531643961766854,0.38422088850782765,-0.0010955511098408866,unchanged
+5gnidd,0.25,0,0.0,3,15967,0.9994923857868021,0.9994923857868021,0.0,0.9989858012170385,0.9989858012170385,0.0,0.8697768301371551,0.8682289990646567,-0.0015478310724984512,0.9611072837727813,0.9588526335567107,-0.0022546502160706305,unchanged
+5gnidd,0.25,0,0.0,4,48464,0.5805717998700455,0.5683318977892621,-0.012239902080783405,1.0,1.0,0.0,0.6223806594118536,0.6183564917563243,-0.004024167655529265,0.712508253549026,0.7119098712446352,-0.0005983823043907988,unchanged
+5gnidd,0.25,0,0.0,5,51668,0.5598073320845598,0.5652250225022503,0.005417690417690468,0.983980625018237,0.983980625018237,0.0,0.8297485288227501,0.8310984610457359,0.0013499322229858501,0.9855423085855849,0.986355190833785,0.0008128822482000864,unchanged
+5gnidd,0.25,0,0.0,6,6258,0.572547299837113,0.5670643528561099,-0.005482946981003156,0.9719887955182073,0.9684873949579832,-0.003501400560224077,0.7072248224060372,0.7056582661257507,-0.0015665562802865685,0.7796420581655481,0.7788430808565037,-0.0007989773090444352,unchanged
+5gnidd,0.25,0,0.0,7,1166,0.9792189581434481,0.978641456582633,-0.0005775015608150413,0.9975550122249388,0.9975550122249388,0.0,0.8763200940745843,0.8753699451417798,-0.0009501489328045398,0.8936535162950258,0.8927958833619211,-0.0008576329331047017,unchanged
+5gnidd,0.25,0,0.0,8,21631,0.9701952723535457,0.9731958762886599,0.0030006039351141434,0.9936842105263158,0.9936842105263158,0.0,0.8201876435777007,0.8112974531570079,-0.008890190420692834,0.9657898386574824,0.9657436087097222,-4.6229947760156875e-05,unchanged
+5gnidd,0.25,0,0.0,9,7529,0.9959349593495934,0.9959349593495934,0.0,1.0,1.0,0.0,0.8698620016089736,0.8716326886457475,0.0017706870367738548,0.9743657856289015,0.975561163501129,0.0011953778722274544,unchanged
+5gnidd,0.25,1,0.0,0,26590,0.45421039634522775,0.4641193036354327,0.009908907290204938,1.0,1.0,0.0,0.6963251183019036,0.6921860065661486,-0.004139111735754986,0.45321549454682214,0.45344114328694995,0.0002256487401278151,unchanged
+5gnidd,0.25,1,0.0,1,15600,0.8915254237288135,0.8915254237288135,0.0,1.0,1.0,0.0,0.7418601393293716,0.7344876958480612,-0.007372443481310387,0.8480769230769231,0.8208974358974359,-0.02717948717948715,unchanged
+5gnidd,0.25,1,0.0,2,21164,0.9887947096334008,0.9887947096334008,0.0,1.0,1.0,0.0,0.8762605174285305,0.8735325605471556,-0.002727956881374949,0.975949725949726,0.9757134757134757,-0.0002362502362502905,unchanged
+5gnidd,0.25,1,0.0,3,25511,0.5,0.5714285714285715,0.07142857142857151,1.0,1.0,0.0,0.6286690326261317,0.62474021042126,-0.003928822204871674,0.761318646858218,0.7618674297361923,0.0005487828779742188,improved
+5gnidd,0.25,1,0.0,4,9968,0.5176902536715621,0.5176902536715621,0.0,1.0,1.0,0.0,0.7380998187344665,0.732750332030492,-0.005349486703974504,0.682784911717496,0.6803772070626003,-0.0024077046548957037,unchanged
+5gnidd,0.25,1,0.0,5,24110,0.15068493150684933,0.15942028985507245,0.008735358348223116,1.0,1.0,0.0,0.4677285241968914,0.4700908451196186,0.0023623209227272013,0.4180837826627955,0.4175031107424305,-0.0005806719203649835,unchanged
+5gnidd,0.25,1,0.0,6,7220,0.9244186046511629,0.9244186046511629,0.0,0.9147727272727273,0.9147727272727273,0.0,0.8795272649707093,0.8708013350727581,-0.008725929897951179,0.8966759002770083,0.8738227146814405,-0.022853185595567815,unchanged
+5gnidd,0.25,1,0.0,7,39377,0.6721281296023565,0.6720273816663547,-0.00010074793600178822,0.9974025974025974,0.9974025974025974,0.0,0.7529499473803684,0.7430173744923633,-0.009932572888005087,0.663534550626,0.6605378774411459,-0.0029966731848540817,unchanged
+5gnidd,0.25,1,0.0,8,27259,0.2835707502374169,0.28270280444193485,-0.0008679457954820502,1.0,1.0,0.0,0.6288051343056038,0.626122535616005,-0.0026825986895988585,0.7458087237242745,0.7455519278036612,-0.00025679592061333967,unchanged
+5gnidd,0.25,1,0.0,9,46378,0.6190476190476191,0.6341463414634146,0.015098722415795574,0.8125,0.8125,0.0,0.7576541807021213,0.753138803537823,-0.004515377164298284,0.9722282116520764,0.9696839018500151,-0.0025443098020613286,unchanged
+5gnidd,0.25,2,0.0,0,25680,0.6181318681318682,0.6181318681318682,0.0,1.0,1.0,0.0,0.6272980982736713,0.6338807440707196,0.006582645797048259,0.6758566978193147,0.6758566978193147,0.0,unchanged
+5gnidd,0.25,2,0.0,1,37469,0.8827838827838828,0.8873015873015873,0.004517704517704502,0.9352941176470588,0.9352941176470588,0.0,0.875873367828011,0.8766380925482959,0.0007647247202849927,0.840721663241613,0.8406949745122635,-2.6688729349522156e-05,unchanged
+5gnidd,0.25,2,0.0,2,8769,0.9625368731563422,0.9625368731563422,0.0,1.0,1.0,0.0,0.8094085650951544,0.8059361017327209,-0.0034724633624334533,0.8022579541566883,0.8064773634393888,0.0042194092827004814,unchanged
+5gnidd,0.25,2,0.0,3,26589,0.8571428571428571,0.8372093023255813,-0.019933554817275767,1.0,1.0,0.0,0.6483212194766437,0.6441355193440161,-0.0041857001326275345,0.4522546917898379,0.45120162473203207,-0.0010530670578058343,unchanged
+5gnidd,0.25,2,0.0,4,14430,0.4665268456375839,0.46393274290848413,-0.0025941027290997676,0.7929292929292929,0.7878787878787878,-0.005050505050505083,0.4794740108344063,0.4781708507446747,-0.001303160089731581,0.4078309078309078,0.4070686070686071,-0.0007623007623007161,unchanged
+5gnidd,0.25,2,0.0,5,44764,0.8272719775688604,0.8320181481636119,0.004746170594751464,0.9176470588235295,0.9176470588235295,0.0,0.6560726383798599,0.6582289531592247,0.002156314779364843,0.5314985256009294,0.5335090697882227,0.002010544187293295,unchanged
+5gnidd,0.25,2,0.0,6,2327,0.9965116279069768,0.9965116279069768,0.0,0.9965116279069768,0.9965116279069768,0.0,0.9158319754659812,0.917676703349529,0.0018447278835478365,0.9428448646325741,0.943704340352385,0.0008594757198109315,unchanged
+5gnidd,0.25,2,0.0,7,58787,0.5513950073421439,0.5536309982037465,0.0022359908616026436,0.9751131221719457,0.9728506787330317,-0.002262443438914019,0.7418357797244598,0.7442211799516925,0.002385400227232659,0.934288192967833,0.9374691683535475,0.0031809753857144907,unchanged
+5gnidd,0.25,2,0.0,8,15019,0.9992129525522824,0.9992129525522824,0.0,0.9988764044943821,0.9988764044943821,0.0,0.8418355327819182,0.8444130692640451,0.0025775364821268854,0.7970570610559957,0.8004527598375392,0.0033956987815434436,unchanged
+5gnidd,0.25,2,0.0,9,9343,0.6399999999999999,0.6399999999999999,0.0,1.0,1.0,0.0,0.5960023609485813,0.5980046801133452,0.0020023191647638994,0.5437225730493418,0.5535695172856684,0.009846944236326594,unchanged
+5gnidd,0.5,0,1.0,0,5719,0.5344946984589947,0.5534193380860204,0.01892463962702573,0.9980392156862745,0.9980392156862745,0.0,0.7937343560428939,0.7956199889478253,0.001885632904931378,0.5980066445182725,0.6013289036544851,0.0033222591362126463,unchanged
+5gnidd,0.5,0,1.0,1,50275,0.9262953525057638,0.9694634172269913,0.04316806472122747,1.0,1.0,0.0,0.7217834392302032,0.7020096073864508,-0.019773831843752365,0.6330382894082546,0.6346295375435107,0.0015912481352561292,improved
+5gnidd,0.5,0,1.0,2,29272,0.7900323027226581,0.8522114347357066,0.06217913201304848,0.9421052631578948,0.9421052631578948,0.0,0.823279670562598,0.816124524768667,-0.00715514579393095,0.9369363213992894,0.9214266192948893,-0.015509702104400058,improved
+5gnidd,0.5,0,1.0,3,41137,0.36627798429895186,0.3088797245787754,-0.05739825972017648,0.9117647058823529,0.9117647058823529,0.0,0.6625036778758502,0.6981128257060828,0.03560914783023261,0.5958626054403579,0.6171816126601356,0.021319007219777752,degraded
+5gnidd,0.5,0,1.0,4,29132,0.7467320261437909,0.8325132130490249,0.08578118690523406,0.9983606557377049,0.9989071038251366,0.0005464480874316502,0.7678751173184121,0.7753499779585092,0.0074748606400971696,0.6180488809556501,0.6155430454483043,-0.002505835507345844,improved
+5gnidd,0.5,0,1.0,5,16968,0.9262510627360263,0.9609720803541029,0.034721017618076555,0.9977973568281938,0.9977973568281938,0.0,0.7542440576096072,0.7773421217044842,0.023098064094877024,0.7514733616218765,0.773986327204149,0.02251296558227245,improved
+5gnidd,0.5,0,1.0,6,27647,0.9265695502619544,0.9261866431027104,-0.0003829071592440503,0.9238626451820227,0.9238626451820227,0.0,0.8872229354418285,0.9170701694549138,0.02984723401308531,0.9714978116974717,0.9762361196513184,0.0047383079538466655,unchanged
+5gnidd,0.5,0,1.0,7,8159,0.8263118994826313,0.7357484620642515,-0.0905634374183798,0.9117647058823529,0.9117647058823529,0.0,0.6403573455692047,0.7018881509455476,0.06153080537634292,0.5370756220125016,0.5677166319401887,0.030641009927687146,degraded
+5gnidd,0.5,0,1.0,8,13584,0.8949672720967532,0.8877277940614248,-0.007239478035328428,0.9384615384615385,0.9384615384615385,0.0,0.6424325008852249,0.7139306163023146,0.07149811541708972,0.6131478209658422,0.6635011778563016,0.05035335689045939,unchanged
+5gnidd,0.5,0,1.0,9,21284,0.9990732159406859,0.9990732159406859,0.0,0.9981481481481481,0.9981481481481481,0.0,0.7656168060197422,0.8859647531966095,0.12034794717686725,0.9905562864123285,0.9976978011651945,0.007141514752865952,unchanged
+5gnidd,0.5,1,1.0,0,35892,0.9986990459670425,0.8230233702913669,-0.17567567567567566,1.0,1.0,0.0,0.7663029928351215,0.7749948243953257,0.008691831560204255,0.896244288420818,0.901315056279951,0.005070767859132963,degraded
+5gnidd,0.5,1,1.0,1,5007,0.9633204633204633,0.8186813186813187,-0.14463914463914462,0.9379562043795621,0.9379562043795621,0.0,0.8429384253186761,0.7992701775496931,-0.04366824776898304,0.8867585380467345,0.8737767126023567,-0.01298182544437787,degraded
+5gnidd,0.5,1,1.0,2,21773,0.9547689282202557,0.9714857428714357,0.016716814651180023,0.9928425357873211,0.9928425357873211,0.0,0.7975560743337922,0.78845847002079,-0.009097604313002217,0.8353465301060947,0.8328204657144169,-0.002526064391677796,unchanged
+5gnidd,0.5,1,1.0,3,36943,0.7476635514018691,0.851063829787234,0.10340027838536492,1.0,1.0,0.0,0.5982514343164378,0.6163240640445933,0.018072629728155487,0.5151178843082587,0.5134666919308123,-0.0016511923774463755,improved
+5gnidd,0.5,1,1.0,4,47087,0.057894736842105256,0.09777777777777778,0.03988304093567253,1.0,1.0,0.0,0.5962439339580969,0.5855008143896245,-0.010743119568472403,0.7007029541062288,0.691337311784569,-0.009365642321659817,improved
+5gnidd,0.5,1,1.0,5,3851,0.9090909090909091,0.9375,0.02840909090909094,0.9375,0.9375,0.0,0.8791902157831661,0.8741577209395658,-0.005032494843600355,0.9504024928589977,0.9407945988055051,-0.009607894053492605,improved
+5gnidd,0.5,1,1.0,6,45750,0.979310344827586,0.9627118644067797,-0.01659848042080636,0.993006993006993,0.993006993006993,0.0,0.6627570270713593,0.7177968261221006,0.0550397990507413,0.5882841530054644,0.5967868852459016,0.008502732240437183,unchanged
+5gnidd,0.5,1,1.0,7,27400,0.970098185063969,0.970098185063969,0.0,0.9977945809703843,0.9977945809703843,0.0,0.8732554749869261,0.8548112853665224,-0.018444189620403728,0.9791605839416059,0.9718248175182482,-0.007335766423357648,unchanged
+5gnidd,0.5,1,1.0,8,11047,0.8048743335872048,0.8275641025641025,0.022689768976897673,0.9313725490196079,0.9313725490196079,0.0,0.8951720353049097,0.8952498845439423,7.784923903253471e-05,0.9516610844573188,0.9352765456685073,-0.01638453878881152,improved
+5gnidd,0.5,1,1.0,9,8427,0.9936708860759493,0.9534883720930232,-0.04018251398292616,1.0,1.0,0.0,0.7209580200426012,0.7124277763940038,-0.008530243648597446,0.8317313397413078,0.7907915035006526,-0.040939836240655114,degraded
+5gnidd,0.5,2,1.0,0,10364,0.0625,0.32142857142857145,0.25892857142857145,1.0,0.8181818181818182,-0.18181818181818177,0.7790490964040604,0.8124727267442877,0.03342363034022733,0.9181783095329988,0.9460632960247009,0.027884986491702013,improved
+5gnidd,0.5,2,1.0,1,74366,0.5432225063938619,0.8399598393574297,0.29673733296356775,1.0,0.9959349593495934,-0.004065040650406582,0.5731920118977858,0.6110237993216254,0.037831787423839636,0.5739854234461985,0.5575935239222225,-0.016391899523976017,improved
+5gnidd,0.5,2,1.0,2,22300,0.8216327283264403,0.7618010426337213,-0.05983168569271902,0.9525129342202513,0.9525129342202513,0.0,0.6102530360051527,0.599765713807183,-0.010487322197969662,0.4416591928251121,0.4420179372197309,0.00035874439461880403,degraded
+5gnidd,0.5,2,1.0,3,10211,0.5959349593495935,1.0,0.4040650406504065,1.0,1.0,0.0,0.8001270328501936,0.8821526176338881,0.08202558478369448,0.8621094897659387,0.8645578297914014,0.002448340025462681,improved
+5gnidd,0.5,2,1.0,4,20659,0.5008064516129033,0.5223787815996684,0.02157232998676517,1.0,1.0,0.0,0.6150616295363164,0.6346139496606714,0.019552320124355038,0.5199186795101408,0.5302289559029962,0.010310276392855422,improved
+5gnidd,0.5,2,1.0,5,13363,0.8141829973213027,0.9613955408753097,0.14721254355400692,1.0,1.0,0.0,0.8418094510297551,0.8824979584017227,0.04068850737196761,0.7961535583327097,0.8065554142033975,0.01040185587068776,improved
+5gnidd,0.5,2,1.0,6,4067,0.8215227272727272,0.9152727272727272,0.09375,1.0,1.0,0.0,0.8707450841900309,0.8610389468215955,-0.009706137368435397,0.8571428571428571,0.849766412589132,-0.007376444553725103,improved
+5gnidd,0.5,2,1.0,7,20924,0.9353669586227725,0.9367816091954023,0.0014146505726297942,0.9042553191489362,0.9042553191489362,0.0,0.902070991788213,0.826329547235495,-0.07574144455271803,0.9935480787612311,0.9874307015866947,-0.00611737717453642,unchanged
+5gnidd,0.5,2,1.0,8,60178,0.8293333333333333,0.8333333333333333,0.0040000000000000036,1.0,1.0,0.0,0.7864084400269984,0.7575747395009161,-0.02883370052608225,0.9910930904981887,0.9773837615075277,-0.013709328990661018,unchanged
+5gnidd,0.5,2,1.0,9,6745,0.8908889241007765,0.9708672086720868,0.07997828457131029,0.9974309569685292,0.9463712267180475,-0.051059730250481716,0.823801233520728,0.8293441306497016,0.005542897128973601,0.9288361749444033,0.894885100074129,-0.03395107487027427,improved
+5gnidd,0.75,0,0.5,0,8392,0.96406593216528,0.9665574550902396,0.002491522924959577,0.9579831932773109,0.9549819927971188,-0.0030012004801920344,0.7525519803108824,0.7478105901015991,-0.004741390209283325,0.6402526215443279,0.6428741658722593,0.00262154432793138,unchanged
+5gnidd,0.75,0,0.5,1,22302,0.941202165225427,0.9412498184292409,4.7653203813946554e-05,0.9263586956521739,0.9263586956521739,0.0,0.7197094523895853,0.7162627765463196,-0.003446675843265634,0.52519953367411,0.5279795534032822,0.0027800197291721895,unchanged
+5gnidd,0.75,0,0.5,2,4449,0.957516339869281,0.9576023391812865,8.599931200548028e-05,0.9605263157894737,0.9605263157894737,0.0,0.8291503247992262,0.8372396118782266,0.008089287079000362,0.8952573612047651,0.8990784445942909,0.0038210833895258034,unchanged
+5gnidd,0.75,0,0.5,3,40148,0.8301259167012591,0.8763328073269594,0.04620689062570027,0.9878669338167477,0.9878669338167477,0.0,0.7754306084907939,0.7965497734403445,0.021119164949550573,0.6591860117565009,0.6622745840390555,0.003088572282554547,improved
+5gnidd,0.75,0,0.5,4,20761,0.8180297883459628,0.8283759074533374,0.010346119107374596,0.9976958525345623,0.9976958525345623,0.0,0.8566532837882197,0.8565822825143011,-7.100127391856947e-05,0.7910505274312413,0.7902316844082655,-0.000818843022975746,unchanged
+5gnidd,0.75,0,0.5,5,37122,0.7981904012588513,0.7618538324420676,-0.03633656881678371,0.9054054054054055,0.9054054054054055,0.0,0.8230345013112859,0.8300206164517117,0.006986115140425797,0.814476590700932,0.8171434728732289,0.002666882172296825,degraded
+5gnidd,0.75,0,0.5,6,34066,0.615909090909091,0.6028761061946903,-0.01303298471440062,1.0,1.0,0.0,0.6684435277862657,0.6570978858235467,-0.011345641962719033,0.7987142605530441,0.7968942640756179,-0.0018199964774261446,unchanged
+5gnidd,0.75,0,0.5,7,22757,0.9523026315789473,0.9508196721311475,-0.0014829594477998587,0.9230769230769231,0.9230769230769231,0.0,0.8202203458150573,0.8501349999086047,0.029914654093547433,0.9323285143032913,0.9403699960451729,0.008041481741881573,unchanged
+5gnidd,0.75,0,0.5,8,27694,0.9515603799185889,0.9318064182194618,-0.01975396169912713,0.9955357142857143,0.9970238095238095,0.0014880952380952328,0.7698856018252216,0.7880392566808838,0.018153654855662227,0.9372427240557522,0.9374232685780314,0.00018054452227922457,unchanged
+5gnidd,0.75,0,0.5,9,25486,0.6656467788543261,0.6839923146930933,0.01834553583876719,0.9982698961937716,0.9982698961937716,0.0,0.6097048579227672,0.6393691849315551,0.02966432700878796,0.5136545554422035,0.5201286981087656,0.006474142666562099,unchanged
+5gnidd,0.75,1,0.5,0,24678,0.5443955604439555,0.7127935980045728,0.16839803756061722,0.9239999999999999,0.9239999999999999,0.0,0.6748195071764608,0.7100234424989598,0.03520393532249899,0.4980144258043602,0.4987032984844801,0.0006888726801199119,improved
+5gnidd,0.75,1,0.5,1,37889,0.0577894327894328,0.07272727272727272,0.014937839937839925,1.0,1.0,0.0,0.6128814319059644,0.6195111566473516,0.006629724741387122,0.6342474068991,0.6340626567077516,-0.00018475019134833737,unchanged
+5gnidd,0.75,1,0.5,2,9750,0.9992483086945627,0.999498997995992,0.00025068930142924106,0.9984977466199298,0.99899849774662,0.0005007511266901243,0.6981053280587833,0.6975887993005325,-0.0005165287582508116,0.804,0.8054358974358974,0.0014358974358973375,unchanged
+5gnidd,0.75,1,0.5,3,28273,0.6571700576651072,0.6582004951570168,0.0010304374919096748,0.9589541338007596,0.9608530528775927,0.0018989190768331499,0.6634466459306729,0.6875494680581039,0.02410282212743098,0.7965903865879107,0.7973685141300888,0.0007781275421780975,unchanged
+5gnidd,0.75,1,0.5,4,23163,0.36669880338002603,0.403711790393013,0.037012987012987,0.9263157894736842,0.9263157894736842,0.0,0.6761318473919267,0.6803473046124073,0.004215457220480667,0.5723783620429133,0.5726373958468247,0.00025903380391145703,improved
+5gnidd,0.75,1,0.5,5,33321,0.8272074098845752,0.888728391582112,0.061520981697536836,0.911578947368421,0.911578947368421,0.0,0.8492588995308481,0.8702507546779032,0.02099185514705515,0.8677410641937517,0.8684313195882476,0.0006902553944958534,improved
+5gnidd,0.75,1,0.5,6,30844,0.5701532510043148,0.7025193798449613,0.13236612884064647,1.0,1.0,0.0,0.729867202070374,0.7560654715107815,0.026198269440407485,0.7726624302943846,0.7725975878614966,-6.484243288806812e-05,improved
+5gnidd,0.75,1,0.5,7,18726,0.9493019982012278,0.9446024651225791,-0.004699533078648788,0.9410569105691057,0.9410569105691057,0.0,0.8111414375751627,0.8468191358182763,0.03567769824311362,0.8352023923955997,0.834935383958133,-0.0002670084374666315,unchanged
+5gnidd,0.75,1,0.5,8,20759,0.965034965034965,0.9737891737891738,0.008754208754208848,1.0,1.0,0.0,0.9181704506988461,0.9210284053529008,0.002857954654054695,0.9496122163880727,0.949949419528879,0.0003372031408063192,unchanged
+5gnidd,0.75,1,0.5,9,15774,0.9956872836986422,0.9956872836986422,0.0,0.9966216216216216,0.9966216216216216,0.0,0.9205558755090844,0.9205839815517529,2.810604266856309e-05,0.9721059972105998,0.9721059972105998,0.0,unchanged
+5gnidd,0.75,2,0.5,0,30869,0.3729455572675912,0.4636986301369863,0.0907530728693951,0.9848484848484849,0.9772727272727273,-0.007575757575757569,0.6481372249101648,0.6675816920608101,0.019444467150645295,0.48867796170915806,0.49055686935112897,0.001878907641970906,improved
+5gnidd,0.75,2,0.5,1,55235,0.9153404067197171,0.9532191945985049,0.037878787878787845,1.0,1.0,0.0,0.8149498232260747,0.8311343595560313,0.016184536329956645,0.9835973567484385,0.981062731963429,-0.0025346247850095427,improved
+5gnidd,0.75,2,0.5,2,21873,0.37836377391116394,0.3924493054927838,0.014085531581619848,1.0,1.0,0.0,0.6356783451236241,0.6407167966698079,0.005038451546183764,0.48104969597220315,0.4830155899967997,0.0019658940245965706,unchanged
+5gnidd,0.75,2,0.5,3,7770,0.9628975265017667,0.9628975265017667,0.0,0.930921052631579,0.930921052631579,0.0,0.9010418178223024,0.9119868973096994,0.01094507948739698,0.9773487773487773,0.9858429858429858,0.008494208494208455,unchanged
+5gnidd,0.75,2,0.5,4,21162,0.8152515723270439,0.8166144200626959,0.0013628477356519841,0.9772727272727273,0.9772727272727273,0.0,0.7644953284759245,0.7805801445013516,0.016084816025427107,0.814667800774974,0.8194405065683773,0.004772705793403276,unchanged
+5gnidd,0.75,2,0.5,5,30481,0.9733161911081619,0.979778870811788,0.0064626797036261285,0.9988636363636363,0.9988636363636363,0.0,0.8808043794681331,0.9045226133222829,0.02371823385414984,0.9648305501787999,0.9659459991470096,0.001115448968209698,unchanged
+5gnidd,0.75,2,0.5,6,25739,0.5179873510499233,0.5199510790231585,0.0019637279732351853,0.9311454311454311,0.9311454311454311,0.0,0.6323145995806274,0.6303763912555371,-0.0019382083250902848,0.541396324643537,0.5411632153541318,-0.00023310928940523112,unchanged
+5gnidd,0.75,2,0.5,7,4581,0.9393065187239944,0.9393065187239944,0.0,0.9583333333333333,0.9583333333333333,0.0,0.9192814934874765,0.9221321422692168,0.0028506487817403503,0.9493560358000437,0.9535036018336608,0.0041475660336171005,unchanged
+5gnidd,0.75,2,0.5,8,7542,0.8026177222807281,0.8164517143755898,0.013833992094861691,0.996551724137931,0.996551724137931,0.0,0.7113836033571382,0.7072279851171424,-0.004155618239995884,0.5350039777247414,0.5328825245293026,-0.0021214531954387716,unchanged
+5gnidd,0.75,2,0.5,9,37925,0.40437943619069516,0.43297343616109685,0.028593999970401685,0.9950980392156863,0.9950980392156863,0.0,0.6409323786438653,0.6777847076048981,0.03685232896103274,0.6937112722478577,0.697139090309822,0.0034278180619643495,improved
@@ -0,0 +1,7 @@
+dataset,alpha,selected_beta,total,improved,degraded,unchanged,mean_delta,median_delta,min_delta,max_delta
+xiiotid,0.25,0.0,30,1,1,28,-0.001517806604221289,0.0,-0.026277262664796908,0.02494573827906854
+xiiotid,0.5,0.75,30,12,1,17,0.03024920134564162,0.012286627106224551,-0.2565947651121153,0.3235930735930736
+xiiotid,0.75,0.5,30,8,2,20,0.01167502563325023,0.010269879504053326,-0.0697526762643042,0.06759906759906764
+5gnidd,0.25,0.0,30,1,0,29,0.002111576617816057,0.0,-0.019933554817275767,0.07142857142857151
+5gnidd,0.5,1.0,30,15,6,9,0.03903400685752253,0.02024848480689545,-0.17567567567567566,0.4040650406504065
+5gnidd,0.75,0.5,30,8,1,21,0.02071412212202052,0.004477101314292853,-0.03633656881678371,0.16839803756061722
@@ -0,0 +1,359 @@
+# Original PROTEAN vs Proposed Beta-Tuned Aggregation: Deep Paired Comparison
+
+Status: **descriptive post-hoc analysis of completed experiments**. No training, no partition
+regeneration, no modification of any result artifact. Every number is derived from the audited
+machine-readable outputs in `results/` and `artifacts/analysis/`, using the reproducible scripts
+`scripts/deep_comparison.py` (derived tables) and `scripts/make_comparison_figures.py`
+(Figures I–K).
+
+Primary comparison:
+
+```
+original_protean            (frozen reproduced baseline)
+        vs
+support_weighted(beta*)     (validation-selected beta per dataset-alpha condition)
+```
+
+This document does **not** substitute `support_weighted(beta=0)` for the baseline anywhere;
+`beta=0` appears only as an explicitly labelled control.
+
+---
+
+## 1. Research question and comparison scope
+
+The question is not whether the proposed weighting is universally better. It is: **where, on the
+completed grid, does the validation-selected `support_weighted` run measurably differ from
+`original_protean`, in which direction, for which clients and classes, and how large are those
+differences relative to seed variability?**
+
+Six dataset-alpha conditions (2 datasets × 3 Dirichlet alpha), three seeds each (18 paired cells, 30
+client-seed observations per condition). All comparisons are on the **test** split, using the same
+partition, split, class mapping, and initialization within each cell.
+
+## 2. Original PROTEAN versus proposed mechanism
+
+Only the **server-side global prototype aggregation** changes:
+
+$$
+\bar{C}_j(\beta) = \frac{\sum_{i \in S_j} N_{ij}^{\beta} C_{ij}}{\sum_{i \in S_j} N_{ij}^{\beta}},
+\qquad
+w_{ij}(\beta) = \frac{N_{ij}^{\beta}}{\sum_{k \in S_j} N_{kj}^{\beta}}
+$$
+
+- `original_protean` uses equal contributions over `S_j` (the reference Eq. 2 behavior under the
+  documented participant rule).
+- `support_weighted(beta*)` reweights each contributing client by its class-specific training support
+  `N_ij`, at the beta selected by pooled validation.
+
+**Unchanged:** local prototype construction, model aggregation, architecture, loss, and the
+nearest-prototype inference rule. **Changed:** only the prototype aggregation weights.
+`N_ij` is class-specific sample support, not prototype quality.
+
+Selected betas (source: `artifacts/analysis/final__selection_strength.csv`), verified:
+
+| Dataset | alpha=0.25 | alpha=0.50 | alpha=0.75 |
+|---|---|---|---|
+| X-IIoTID | 0.00 | 0.75 | 0.50 |
+| 5G-NIDD | 0.00 | 1.00 | 0.50 |
+
+## 3. Pairing and evaluation methodology
+
+- **Pairing verified** for all 18 cells: same dataset, alpha, seed, and **identical
+  `manifest_content_hash`** between the `original_protean` and `support_weighted` runs
+  (`scripts/deep_comparison.py` raises on any mismatch).
+- `Delta = proposed − original` everywhere. Percentage points (pp) are never conflated with relative
+  percent.
+- Per-client test metrics come from each run's `per_client.csv` (rare-class macro-F1 and recall are
+  computed on each client's own evaluation partition using that client's own rare-class ids;
+  all 10 clients are eligible in every condition). Per-class metrics come from `test_per_class.csv`;
+  confusion matrices from `test_confusion_matrix.json`.
+- **Effectively-unchanged band = ±0.02** for paired per-client deltas, chosen to reflect the observed
+  `original_protean` vs `support_weighted(beta=0)` residual divergence (the two "selected beta = 0"
+  conditions). Raw deltas are stored in `deepcmp__per_client.csv` so any other threshold can be
+  applied.
+- Three seeds do not support inferential significance testing; all statements are descriptive.
+
+Derived tables: `deepcmp__global_per_seed.csv`, `deepcmp__global_summary.csv`,
+`deepcmp__per_client.csv`, `deepcmp__per_client_by_condition.csv`, `deepcmp__per_class.csv`,
+`deepcmp__confusion_delta.csv`, `deepcmp__client_support_join.csv`.
+
+## 4. Six-condition global metric comparison
+
+Test metrics, mean over seeds `{0,1,2}`; `Δ = proposed − original` (source:
+`deepcmp__global_summary.csv`).
+
+| Dataset | alpha | beta* | rare-F1 orig | rare-F1 prop | **Δ rare-F1** | Δ rare-recall | Δ macro-F1 | Δ weighted-F1 | Δ accuracy |
+|---|---|---|---|---|---|---|---|---|---|
+| X-IIoTID | 0.25 | 0.00 | 0.6328 | 0.6321 | **−0.0007** | −0.0007 | −0.0007 | −0.0008 | −0.0014 |
+| X-IIoTID | 0.50 | 0.75 | 0.6425 | 0.6655 | **+0.0230** | −0.0144 | +0.0222 | +0.0089 | +0.0069 |
+| X-IIoTID | 0.75 | 0.50 | 0.6039 | 0.6052 | **+0.0014** | −0.0077 | +0.0058 | +0.0061 | +0.0075 |
+| 5G-NIDD | 0.25 | 0.00 | 0.8423 | 0.8412 | **−0.0010** | +0.0007 | −0.0028 | −0.0010 | −0.0007 |
+| 5G-NIDD | 0.50 | 1.00 | 0.8805 | 0.8720 | **−0.0086** | −0.0058 | −0.0038 | −0.0005 | −0.0007 |
+| 5G-NIDD | 0.75 | 0.50 | 0.8799 | 0.8967 | **+0.0168** | −0.0015 | +0.0117 | +0.0014 | +0.0013 |
+
+Per-seed rare-class macro-F1 deltas (seed 0 / 1 / 2), showing the seed spread:
+
+| Dataset | alpha | beta* | Δ seed 0 | Δ seed 1 | Δ seed 2 |
+|---|---|---|---|---|---|
+| X-IIoTID | 0.25 | 0.00 | +0.0016 | −0.0020 | −0.0016 |
+| X-IIoTID | 0.50 | 0.75 | +0.0266 | +0.0163 | +0.0261 |
+| X-IIoTID | 0.75 | 0.50 | −0.0097 | +0.0101 | +0.0036 |
+| 5G-NIDD | 0.25 | 0.00 | −0.0050 | −0.0010 | +0.0029 |
+| 5G-NIDD | 0.50 | 1.00 | +0.0090 | −0.0811 | +0.0464 |
+| 5G-NIDD | 0.75 | 0.50 | −0.0011 | +0.0450 | +0.0067 |
+
+Interpretation:
+
+- **Two conditions are numerically null**: both `alpha=0.25` conditions selected `beta=0`, so the
+  selected run is the equal-weight `support_weighted` run; their small deltas are floating-point
+  divergence, not a weighting effect.
+- Among the **four genuine weighting conditions (beta>0)**, rare-class macro-F1 improved in three
+  (X-IIoTID a0.50 +0.023, X-IIoTID a0.75 +0.001, 5G-NIDD a0.75 +0.017) and degraded in one
+  (5G-NIDD a0.50 −0.009).
+- 5G-NIDD a0.50 is dominated by **seed variance** (−0.081 at seed 1 versus +0.046 at seed 2); its
+  pooled mean is not a reliable condition-level statement.
+- **Rare-class recall drops in five of six conditions** (and in **all four** beta>0 conditions)
+  even when rare-class macro-F1 rises: the F1 gains are largely precision-driven (see §7).
+- Overall macro-F1 moves **in the same direction** as rare-class F1 in every condition; where rare-F1
+  improves, overall macro-F1 also improves (X-IIoTID a0.50 +0.022, 5G-NIDD a0.75 +0.012). There is
+  no condition with a rare-class gain accompanied by an overall-metric collapse.
+
+## 5. Rare-class performance trade-offs
+
+The recurring pattern is a **precision–recall trade-off**, not a uniform improvement:
+
+- X-IIoTID a0.50: rare-F1 +0.0230 while rare-recall −0.0144 → rare-class precision rose.
+- X-IIoTID a0.75: rare-F1 +0.0014 while rare-recall −0.0077.
+- 5G-NIDD a0.75: rare-F1 +0.0168 while rare-recall −0.0015.
+- 5G-NIDD a0.50: rare-F1 −0.0086 and rare-recall −0.0058 (both down; not a trade-off but a
+  regression, and seed-dominated).
+
+So the weighting can raise the rare-class **balance** metric by suppressing false positives on rare
+classes while giving up a little rare-class coverage.
+
+## 6. Paired per-client results (test, vs original_protean)
+
+Figure I: `artifacts/figures/final/figI_per_client_original_vs_proposed.png`.
+Per-observation table: `deepcmp__per_client.csv` (30 observations per condition).
+
+| Condition | beta* | improved | degraded | unchanged | mean Δ | min Δ | max Δ |
+|---|---|---|---|---|---|---|---|
+| X-IIoTID a0.25 | 0.00 | 1 | 1 | 28 | −0.0015 | −0.026 | +0.025 |
+| X-IIoTID a0.50 | 0.75 | 12 | 1 | 17 | +0.0302 | −0.257 | +0.324 |
+| X-IIoTID a0.75 | 0.50 | 8 | 2 | 20 | +0.0117 | −0.070 | +0.068 |
+| 5G-NIDD a0.25 | 0.00 | 1 | 0 | 29 | +0.0021 | −0.020 | +0.071 |
+| 5G-NIDD a0.50 | 1.00 | 15 | 6 | 9 | +0.0390 | −0.176 | +0.404 |
+| 5G-NIDD a0.75 | 0.50 | 8 | 1 | 21 | +0.0207 | −0.036 | +0.168 |
+
+Counts use the ±0.02 effectively-unchanged band on per-client test rare-class macro-F1.
+
+- At the two `beta=0` conditions the counts collapse to essentially "unchanged" (28/30 and 29/30),
+  confirming that the beta=0 run is numerically the equal-weight control.
+- At beta>0 conditions the response is **heterogeneous**: X-IIoTID a0.50 improves a majority
+  (12/30) with only 1 clear degradation but contains the single largest loss (−0.257); 5G-NIDD a0.50
+  improves 15/30 but degrades 6/30.
+- Client deltas are not small in the tails: individual clients move by ±0.17–0.40 at the strongest
+  conditions. A condition-level mean therefore hides material client-level redistribution.
+- Client ids are per-seed realizations; id `k` is not the same entity across seeds and is never
+  pooled.
+
+## 7. Per-class precision/recall/F1 changes (beta>0 conditions)
+
+Figure J: `artifacts/figures/final/figJ_per_class_original_vs_proposed.png`.
+Source: `deepcmp__per_class.csv` (per seed) aggregated as the mean delta over the three seeds; class
+support is the mean test support over seeds. X-IIoTID and 5G-NIDD labels are never mixed.
+
+X-IIoTID, alpha=0.50 (beta=0.75):
+
+| Class | Δ F1 | Δ recall | Δ precision | support~ |
+|---|---|---|---|---|
+| Tampering | **+0.144** | −0.009 | +0.212 | 1,024 |
+| Reconnaissance | +0.054 | +0.086 | −0.003 | 25,474 |
+| Lateral_movement | +0.037 | +0.059 | +0.004 | 6,319 |
+| Weaponization | +0.019 | −0.029 | +0.065 | 12,884 |
+| Normal | −0.006 | −0.012 | −0.000 | 83,651 |
+| C&C | **−0.056** | −0.105 | −0.034 | 573 |
+
+X-IIoTID, alpha=0.75 (beta=0.50): largest gains Reconnaissance (+0.033) and Tampering (+0.010);
+largest loss Exfiltration (−0.013).
+
+5G-NIDD, alpha=0.50 (beta=1.00): largest gain TCPConnectScan (+0.034, precision-driven); largest
+losses SYNFlood (−0.043, precision −0.073) and SlowrateDoS (−0.013). 5G-NIDD, alpha=0.75 (beta=0.50):
+largest gain ICMPFlood (+0.097, precision +0.102 on a very small support of ~231); losses SYNFlood
+(−0.009).
+
+Points to note:
+
+- Gains are **not confined to rare classes**: at X-IIoTID a0.50 the largest gain is on the
+  low-support class **Tampering**, but a large gain also appears on the higher-support class
+  **Reconnaissance** (recall-driven).
+- The largest per-class **degradations** are on low-support attack classes with volatile F1:
+  **C&C** at X-IIoTID a0.50 and **SYNFlood** at 5G-NIDD a0.50. Both are precision losses at very
+  small support (~573 and ~1,944).
+- **Benign detection is essentially unchanged** (Δ F1 +0.0009 to +0.0019; recall within ±0.002),
+  so the weighting does not trade benign detection for rare-class gains.
+- Increased rare-class recall does **not** systematically accompany increased false positives:
+  X-IIoTID Reconnaissance recall rose +0.086 with precision −0.003, i.e. almost free recall.
+  Conversely, Tampering F1 rose via precision, not recall.
+
+## 8. Confusion-matrix difference analysis
+
+Figure K shows the paired difference for **5G-NIDD, alpha=0.75, seed=0** (the same pair as Figure H):
+`artifacts/figures/final/figK_confusion_delta_5gnidd_a0.75_seed0.png`. Rows = true class, columns =
+predicted class; both matrices have identical class order and 243,177 test samples. Source:
+`deepcmp__confusion_delta.csv`.
+
+Largest raw-count changes for this pair (proposed − original):
+
+| Change | true → predicted | original | proposed | Δ count | Δ row-rate |
+|---|---|---|---|---|---|
+| Reduced error | HTTPFlood → SlowrateDoS | 759 | 212 | **−547** | −1.94 pp |
+| Reduced error | Benign → TCPConnectScan | 2,015 | 1,495 | **−520** | −0.54 pp |
+| Increased error | SlowrateDoS → HTTPFlood | 1,079 | 1,308 | **+229** | +1.57 pp |
+| Increased error | Benign → UDPFlood | 56,899 | 57,054 | **+155** | +0.16 pp |
+| Diagonal | HTTPFlood → HTTPFlood | 27,399 | 27,946 | +547 | +1.94 pp |
+| Diagonal | Benign → Benign | 36,263 | 36,551 | +288 | +0.30 pp |
+| Diagonal | SlowrateDoS → SlowrateDoS | 13,480 | 13,241 | −239 | −1.63 pp |
+
+- The dominant residual error, **Benign → UDPFlood (~57k samples)**, is essentially unchanged; the
+  proposed weighting neither fixes nor meaningfully worsens it. The +155 cell must not be described
+  as an improvement.
+- The main benefit here is a **net reduction of the HTTPFlood ↔ SlowrateDoS confusion**: the
+  HTTPFlood → SlowrateDoS cell lost 547 samples while the reverse gained 229, and HTTPFlood diagonal
+  gained 547 — consistent with higher HTTPFlood precision.
+- The main cost here is a **small loss of SlowrateDoS diagonal** (−239) with a +229 shift into
+  SlowrateDoS → HTTPFlood.
+- Across all four beta>0 conditions (summed raw counts), the largest **reduced** misclassifications
+  are X-IIoTID `Reconnaissance → Exploitation` (−6,773), `Reconnaissance → Weaponization` (−3,398),
+  `Lateral_movement → Tampering` (−1,699) and 5G-NIDD `Benign → TCPConnectScan` (−1,484); the largest
+  **increased** ones are X-IIoTID `Normal → C&C` (+2,546) and `Weaponization → Exploitation` (+1,286).
+  Raw counts are support-sensitive, so these are descriptive; the row-normalized view (Figure K,
+  right panel) is the support-fair comparison.
+
+## 9. Support imbalance versus client outcomes
+
+Join: per-client test delta (beta>0 conditions only) against the client's training support from
+`rare_unseen.csv` (`deepcmp__client_support_join.csv`, n = 120 client-seed observations).
+
+| Statistic | Value |
+|---|---|
+| Pearson(Δ rare-F1, rare-class support fraction) | −0.092 |
+| Pearson(Δ rare-F1, total client samples) | −0.009 |
+| Low rare-support quartile (≤72 rare-class samples) mean Δ | **+0.041** |
+| High rare-support quartile (≥290) mean Δ | **+0.008** |
+| Mean Δ for clients with 0 missing classes (n=101) | +0.022 |
+| Mean Δ for clients with exactly 1 missing class (n=16) | +0.045 |
+
+- There is **no evidence that low-support clients are disproportionately harmed**; the linear
+  association is near zero and the low-support quartile improved, on average, at least as much as the
+  high-support quartile.
+- Clients with one missing class improved more on average, but n is small.
+- **Limitation:** this is a descriptive association across 120 noisy observations, not a causal test;
+  the beta>0 conditions confound `beta`, dataset, and alpha, and no significance test is valid with
+  three seeds.
+
+## 10. Prototype displacement and interpretation
+
+From `final__prototype_displacement_by_beta.csv` (fixed round-0 snapshots, identical local prototypes
+and `N_ij`, no retraining), mean `‖Cbar_j(1) − Cbar_j(0)‖`:
+
+| Dataset | alpha=0.25 | alpha=0.50 | alpha=0.75 |
+|---|---|---|---|
+| X-IIoTID | 3.422 | 2.099 | 1.305 |
+| 5G-NIDD | 2.919 | 2.340 | 1.385 |
+
+- Displacement is monotone in beta and largest where realized support imbalance is largest.
+- The observed per-class/per-client detection changes are **consistent with** the mechanism chain
+  support imbalance → weight redistribution → prototype displacement → detection change, but the
+  experiment **does not isolate** the displacement→detection step (training is end-to-end).
+  Fixed-snapshot displacement must not be read as detection quality.
+
+## 11. Conditions where the proposed method helps
+
+- **X-IIoTID α=0.50 (beta=0.75)**: the clearest case. Rare-class macro-F1 +0.0230, overall macro-F1
+  +0.0222, accuracy +0.0069; 12/30 client observations improved (1 degraded). Gains driven by
+  Reconnaissance recall and Tampering precision; rare-class recall traded down slightly.
+- **5G-NIDD α=0.75 (beta=0.50)**: rare-class macro-F1 +0.0168, macro-F1 +0.0117; 8/30 improved,
+  1 degraded.
+- **X-IIoTID α=0.75 (beta=0.50)**: marginal rare-F1 gain (+0.0014) with broader macro-F1/accuracy
+  gains (+0.0058/+0.0075).
+
+## 12. Conditions where the proposed method hurts
+
+- **5G-NIDD α=0.50 (beta=1.00)**: rare-class macro-F1 −0.0086 and macro-F1 −0.0038, seed-dominated
+  (−0.081 at seed 1). 6/30 client observations degraded. The clearest case against stronger weighting.
+- **Both α=0.25 conditions**: the selected beta is 0 (equal weighting), so there is no weighting
+  benefit; the small deltas are numerical. At `beta=1` these conditions show the largest per-client
+  degradations, which is why validation did not select a nonzero beta there.
+- **Specific classes**: X-IIoTID **C&C** (−0.056 F1) and 5G-NIDD **SYNFlood** (−0.043 F1) degrade at
+  the strongest `beta`, both low-support classes, and the 5G-NIDD α=0.75 seed-0 confusion shows a
+  small SlowrateDoS recall loss.
+- **No condition** shows a large, consistent, cross-seed rare-class improvement; the largest positive
+  condition mean is +0.023 and its strongest seed-rate change is within the seed spread.
+
+## 13. Effect sizes and practical relevance
+
+- Genuine rare-class macro-F1 effects are **≤ ~0.023** in magnitude, i.e. **≤ 2.3 percentage
+  points**, and the largest is roughly the size of the between-seed standard deviation at that
+  condition. They are small relative to seed noise.
+- The clearest overall-metric movement is X-IIoTID α=0.50 (macro-F1 +0.022, accuracy +0.007); the
+  other positive conditions move by ≤ ~0.012 macro-F1.
+- Client-level redistribution is larger than the condition-level effect: individual clients move by
+  up to −0.26 / +0.32 (X-IIoTID α=0.50) even though the mean is +0.03.
+- Practically, the weighting is **not a reliable blanket improvement**; it is a small, condition- and
+  client-dependent adjustment whose sign is not predictable from alpha alone.
+
+## 14. Threats to validity
+
+- **Only three seeds**; no inferential test is valid. Condition means are dominated by seed variance
+  at 5G-NIDD α=0.50.
+- **Two `beta=0` conditions are numerical, not weighting effects**; they are retained for
+  completeness and labelled.
+- **`S_j` is a reproduction assumption** (no official PROTEAN code). All comparisons are conditional
+  on both modes sharing it.
+- **Rare-class composition differs by dataset**; absolute cross-dataset levels are not comparable.
+- **Per-class metrics on tiny supports** (C&C ~573, ICMPFlood ~231, SYNFlood ~1,944) are volatile.
+- **Confusion-matrix raw counts** are support-sensitive; use row-normalized rates for support-fair
+  comparison.
+- Client ids are per-seed realizations and are not pooled across seeds.
+- `original_protean` is never substituted by `beta=0`.
+
+## 15. Evidence-backed scientific conclusions
+
+1. Over the six conditions, the validation-selected beta produces **small, mixed-sign** test
+   differences from `original_protean`; it is not a universal improvement.
+2. Among the four genuine `beta>0` conditions, rare-class macro-F1 improved in **three** and degraded
+   in **one**; the largest gain is ~2.3 pp and is within seed variability.
+3. When rare-class macro-F1 rises, **overall macro-F1 rises with it**, but **rare-class recall
+   usually falls** — the gains are precision-driven.
+4. Benefits and harms are **client-specific and class-specific**; within a condition some clients
+   gain and others lose, with individual changes larger than the condition mean.
+5. The dominant classification error (Benign → UDPFlood) is **unchanged** by the weighting. The
+   measurable benefit at the analysed pair is a reduced HTTPFlood ↔ SlowrateDoS confusion.
+6. Low-support clients are **not** systematically disadvantaged in the observed data, but the
+   evidence is descriptive only.
+7. Where weighting does not help (both α=0.25 conditions), validation correctly selects `beta=0`.
+
+**Final question — what measurable benefit does beta-controlled prototype aggregation provide over
+original PROTEAN, under which conditions, and at what cost?**
+The evidence supports a **modest, condition-dependent benefit**: at X-IIoTID α=0.50 a rare-class
+macro-F1 gain of about **+0.023** with a matching overall macro-F1 gain, and smaller gains at X-IIoTID
+α=0.75 and 5G-NIDD α=0.75. The cost is a small **rare-class recall reduction** (precision–recall
+trade-off), some **client- and class-level regressions** (e.g. X-IIoTID C&C, 5G-NIDD SYNFlood), and a
+clear **regression at 5G-NIDD α=0.50**. At α=0.25 the best tested beta is 0 (no benefit). The
+evidence does **not** support a consistent advantage, and no claim of statistical superiority is
+made.
+
+---
+
+### Source map
+
+| Claim group | Source |
+|---|---|
+| Pairing, global/per-client/per-class/confusion deltas | `scripts/deep_comparison.py` → `artifacts/analysis/deepcmp__*.csv` |
+| Selected betas | `artifacts/analysis/final__selection_strength.csv` |
+| Global test means/stds | `artifacts/analysis/final__six_condition_master.csv` |
+| Selected-vs-original descriptive deltas | `artifacts/analysis/final__selected_vs_original.csv` |
+| Prototype displacement | `artifacts/analysis/final__prototype_displacement_by_beta.csv` |
+| Per-client / per-class / confusion raw | `results/<run>/per_client.csv`, `test_per_class.csv`, `test_confusion_matrix.json` |
+| Figures I, J, K | `scripts/make_comparison_figures.py` → `artifacts/figures/final/figI/figJ/figK*.{png,pdf}` |
@@ -0,0 +1,239 @@
+#!/usr/bin/env python3
+"""Deep paired comparison: original_protean vs support_weighted(selected beta).
+
+Reads only completed experimental artifacts under ``results/`` and
+``artifacts/analysis/`` and writes derived comparison tables under
+``artifacts/analysis/``. No training, no result mutation.
+
+For every (dataset, alpha, seed) the ``original_protean`` run and the
+``support_weighted`` run at the validation-selected beta are matched by
+dataset, alpha, seed, and manifest content hash, and compared on test
+metrics (global, per-client, per-class, confusion matrices).
+
+Run:
+    .venv/bin/python scripts/deep_comparison.py
+"""
+
+from __future__ import annotations
+
+import csv
+import json
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parents[1]
+RESULTS = ROOT / "results"
+ANALYSIS = ROOT / "artifacts" / "analysis"
+
+DATASETS = ["xiiotid", "5gnidd"]
+ALPHAS = [0.25, 0.50, 0.75]
+SEEDS = [0, 1, 2]
+# Validation-selected beta per condition (source: final__selection_strength.csv).
+SELECTED = {("xiiotid", 0.25): 0.0, ("xiiotid", 0.5): 0.75, ("xiiotid", 0.75): 0.5,
+            ("5gnidd", 0.25): 0.0, ("5gnidd", 0.5): 1.0, ("5gnidd", 0.75): 0.5}
+GLOBAL_METRICS = ["rare_class_macro_f1", "rare_class_recall", "macro_f1",
+                  "weighted_f1", "accuracy"]
+# "Effectively unchanged" band for paired per-client deltas. Chosen to reflect
+# the observed original_protean vs support_weighted(beta=0) residual divergence
+# (the beta=0 conditions are only semantically equivalent to the original under
+# the shared S_j rule, so their per-client residuals are numerical, not a
+# weighting effect). Reported alongside the raw deltas so a reader can apply a
+# different threshold.
+TOL = 0.02
+
+
+def run_dir(ds: str, alpha: float, seed: int, mode: str, beta: float | None = None) -> Path:
+    if mode == "original_protean":
+        return RESULTS / f"{ds}__alpha{alpha:g}__seed{seed}__original_protean"
+    return RESULTS / f"{ds}__alpha{alpha:g}__seed{seed}__support_weighted__beta{beta:g}"
+
+
+def read_json(p: Path):
+    return json.loads(p.read_text())
+
+
+def read_csv(p: Path) -> list[dict]:
+    with p.open() as fh:
+        return list(csv.DictReader(fh))
+
+
+def verify_pairing(ds, alpha, seed, beta) -> tuple[bool, dict]:
+    o = read_json(run_dir(ds, alpha, seed, "original_protean") / "metadata.json")
+    s = read_json(run_dir(ds, alpha, seed, "support_weighted", beta) / "metadata.json")
+    ok = (o["dataset"] == s["dataset"] == ds
+          and abs(float(o["alpha"]) - alpha) < 1e-9
+          and abs(float(s["alpha"]) - alpha) < 1e-9
+          and int(o["seed"]) == int(s["seed"]) == seed
+          and o["manifest_content_hash"] == s["manifest_content_hash"]
+          and abs(float(s["beta"]) - beta) < 1e-9)
+    return ok, o
+
+
+def main() -> None:
+    ANALYSIS.mkdir(parents=True, exist_ok=True)
+    global_rows, client_rows, class_rows, cm_rows = [], [], [], []
+    conditions = []
+
+    for ds in DATASETS:
+        for alpha in ALPHAS:
+            beta = SELECTED[(ds, alpha)]
+            cond = {"dataset": ds, "alpha": alpha, "selected_beta": beta}
+            conditions.append(cond)
+            for seed in SEEDS:
+                ok, meta = verify_pairing(ds, alpha, seed, beta)
+                if not ok:
+                    raise SystemExit(f"pairing failed for {ds} a{alpha} s{seed}")
+                odir = run_dir(ds, alpha, seed, "original_protean")
+                sdir = run_dir(ds, alpha, seed, "support_weighted", beta)
+
+                og = read_json(odir / "test_global_metrics.json")
+                sg = read_json(sdir / "test_global_metrics.json")
+                for m in GLOBAL_METRICS:
+                    global_rows.append({
+                        "dataset": ds, "alpha": alpha, "seed": seed,
+                        "selected_beta": beta, "metric": m,
+                        "original": og[m], "proposed": sg[m],
+                        "delta": sg[m] - og[m]})
+
+                opc = {int(r["client_id"]): r for r in read_csv(odir / "per_client.csv")}
+                spc = {int(r["client_id"]): r for r in read_csv(sdir / "per_client.csv")}
+                assert set(opc) == set(spc), "client id mismatch"
+                for cid in sorted(opc):
+                    o, s = opc[cid], spc[cid]
+                    dr = float(s["rare_class_macro_f1"]) - float(o["rare_class_macro_f1"])
+                    dm = float(s["macro_f1"]) - float(o["macro_f1"])
+                    da = float(s["accuracy"]) - float(o["accuracy"])
+                    dre = float(s["rare_class_recall"]) - float(o["rare_class_recall"])
+                    status = ("unchanged" if abs(dr) <= TOL
+                              else ("improved" if dr > 0 else "degraded"))
+                    client_rows.append({
+                        "dataset": ds, "alpha": alpha, "seed": seed,
+                        "selected_beta": beta, "client_id": cid,
+                        "n_samples": int(o["n_samples"]),
+                        "orig_rare_f1": o["rare_class_macro_f1"],
+                        "prop_rare_f1": s["rare_class_macro_f1"],
+                        "delta_rare_f1": dr,
+                        "orig_rare_recall": o["rare_class_recall"],
+                        "prop_rare_recall": s["rare_class_recall"],
+                        "delta_rare_recall": dre,
+                        "orig_macro_f1": o["macro_f1"], "prop_macro_f1": s["macro_f1"],
+                        "delta_macro_f1": dm,
+                        "orig_accuracy": o["accuracy"], "prop_accuracy": s["accuracy"],
+                        "delta_accuracy": da,
+                        "status": status})
+
+                opk = {int(r["class_id"]): r for r in read_csv(odir / "test_per_class.csv")}
+                spk = {int(r["class_id"]): r for r in read_csv(sdir / "test_per_class.csv")}
+                for cl in sorted(opk):
+                    o, s = opk[cl], spk[cl]
+                    class_rows.append({
+                        "dataset": ds, "alpha": alpha, "seed": seed,
+                        "class_id": cl, "class_name": o["class_name"],
+                        "support": int(o["support"]),
+                        "orig_precision": o["precision"], "prop_precision": s["precision"],
+                        "delta_precision": float(s["precision"]) - float(o["precision"]),
+                        "orig_recall": o["recall"], "prop_recall": s["recall"],
+                        "delta_recall": float(s["recall"]) - float(o["recall"]),
+                        "orig_f1": o["f1"], "prop_f1": s["f1"],
+                        "delta_f1": float(s["f1"]) - float(o["f1"])})
+
+                ocm = read_json(odir / "test_confusion_matrix.json")
+                scm = read_json(sdir / "test_confusion_matrix.json")
+                names = meta["class_names"]
+                for i, t in enumerate(names):
+                    row_o = sum(ocm[i])
+                    row_s = sum(scm[i])
+                    for j, p in enumerate(names):
+                        cm_rows.append({
+                            "dataset": ds, "alpha": alpha, "seed": seed,
+                            "true_class": t, "pred_class": p,
+                            "orig_count": ocm[i][j], "prop_count": scm[i][j],
+                            "delta": scm[i][j] - ocm[i][j],
+                            "orig_row_rate": (ocm[i][j] / row_o) if row_o else 0.0,
+                            "prop_row_rate": (scm[i][j] / row_s) if row_s else 0.0,
+                            "delta_row_rate": ((scm[i][j] / row_s) - (ocm[i][j] / row_o))
+                            if row_o and row_s else 0.0})
+
+    _write(ANALYSIS / "deepcmp__global_per_seed.csv", global_rows)
+    _write(ANALYSIS / "deepcmp__per_client.csv", client_rows)
+    _write(ANALYSIS / "deepcmp__per_class.csv", class_rows)
+    _write(ANALYSIS / "deepcmp__confusion_delta.csv", cm_rows)
+
+    # Summaries -------------------------------------------------------------
+    summary = []
+    for c in conditions:
+        ds, alpha, beta = c["dataset"], c["alpha"], c["selected_beta"]
+        for m in GLOBAL_METRICS:
+            rows = [r for r in global_rows
+                    if r["dataset"] == ds and r["alpha"] == alpha and r["metric"] == m]
+            orig = [float(r["original"]) for r in rows]
+            prop = [float(r["proposed"]) for r in rows]
+            dlt = [float(r["delta"]) for r in rows]
+            summary.append({
+                "dataset": ds, "alpha": alpha, "selected_beta": beta, "metric": m,
+                "orig_mean": _mean(orig), "orig_std": _std(orig),
+                "prop_mean": _mean(prop), "prop_std": _std(prop),
+                "delta_mean": _mean(dlt), "delta_std": _std(dlt),
+                "per_seed_deltas": "|".join(f"{d:.6f}" for d in dlt), "n": len(rows)})
+    _write(ANALYSIS / "deepcmp__global_summary.csv", summary)
+
+    for c in conditions:
+        ds, alpha = c["dataset"], c["alpha"]
+        rows = [r for r in client_rows if r["dataset"] == ds and r["alpha"] == alpha]
+        dlt = [float(r["delta_rare_f1"]) for r in rows]
+        st = [r["status"] for r in rows]
+        c["improved"] = st.count("improved")
+        c["degraded"] = st.count("degraded")
+        c["unchanged"] = st.count("unchanged")
+        c["total"] = len(rows)
+        c["mean_delta"] = _mean(dlt)
+        c["median_delta"] = _median(dlt)
+        c["min_delta"] = min(dlt)
+        c["max_delta"] = max(dlt)
+    _write(ANALYSIS / "deepcmp__per_client_by_condition.csv", [
+        {k: c[k] for k in ("dataset", "alpha", "selected_beta", "total",
+                           "improved", "degraded", "unchanged",
+                           "mean_delta", "median_delta", "min_delta", "max_delta")}
+        for c in conditions])
+
+    print("=== GLOBAL SUMMARY (delta = proposed - original) ===")
+    for r in summary:
+        print(f"  {r['dataset']:8} a{r['alpha']:<4} beta={r['selected_beta']:<4} "
+              f"{r['metric']:20} orig={float(r['orig_mean']):.4f} "
+              f"prop={float(r['prop_mean']):.4f} delta={float(r['delta_mean']):+.4f} "
+              f"per-seed=[{r['per_seed_deltas']}]")
+    print("=== PER-CLIENT COUNTS (rare-class F1) ===")
+    for c in conditions:
+        print(f"  {c['dataset']:8} a{c['alpha']:<4} beta={c['selected_beta']:<4} "
+              f"improved={c['improved']} degraded={c['degraded']} "
+              f"unchanged={c['unchanged']} (of {c['total']}) "
+              f"meanDelta={c['mean_delta']:+.4f} min={c['min_delta']:+.4f} "
+              f"max={c['max_delta']:+.4f}")
+    print("wrote deepcmp__* derived tables to", ANALYSIS)
+
+
+def _mean(v):
+    return sum(v) / len(v) if v else 0.0
+
+
+def _std(v):
+    if len(v) < 2:
+        return 0.0
+    m = _mean(v)
+    return (sum((x - m) ** 2 for x in v) / (len(v) - 1)) ** 0.5
+
+
+def _median(v):
+    s = sorted(v)
+    n = len(s)
+    return s[n // 2] if n % 2 else (s[n // 2 - 1] + s[n // 2]) / 2
+
+
+def _write(path: Path, rows: list[dict]) -> None:
+    with path.open("w", newline="") as fh:
+        w = csv.DictWriter(fh, fieldnames=list(rows[0].keys()))
+        w.writeheader()
+        w.writerows(rows)
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,229 @@
+#!/usr/bin/env python3
+"""Figures I, J, K: original_protean vs support_weighted(selected beta).
+
+Figure I  paired per-client rare-class F1 deltas (test), vs original_protean.
+Figure J  per-class test F1 change, beta>0 conditions.
+Figure K  test confusion-matrix difference (5G-NIDD, alpha=0.75, seed=0).
+
+Reads derived tables written by ``scripts/deep_comparison.py`` and the raw
+per-run confusion matrices. Writes PNG (220 dpi) and PDF to
+``artifacts/figures/final``. No training, no result mutation.
+
+Run:
+    .venv/bin/python scripts/make_comparison_figures.py
+"""
+
+from __future__ import annotations
+
+import csv
+import json
+from collections import defaultdict
+from pathlib import Path
+import statistics
+
+import matplotlib
+
+matplotlib.use("Agg")
+import matplotlib.pyplot as plt
+import numpy as np
+from matplotlib.colors import TwoSlopeNorm
+
+ROOT = Path(__file__).resolve().parents[1]
+ANALYSIS = ROOT / "artifacts" / "analysis"
+RESULTS = ROOT / "results"
+OUT = ROOT / "artifacts" / "figures" / "final"
+
+CHARCOAL = "#3A3A3A"
+TEAL = "#0E7C7B"
+RED = "#C0392B"
+GREY = "#9A9A9A"
+DATASETS = ["xiiotid", "5gnidd"]
+DATASET_NAME = {"xiiotid": "X-IIoTID", "5gnidd": "5G-NIDD"}
+ALPHAS = [0.25, 0.50, 0.75]
+TOL = 0.02
+SEED_MARKERS = {0: "o", 1: "s", 2: "^"}
+
+
+def _style():
+    plt.rcParams.update({
+        "figure.facecolor": "white", "axes.facecolor": "white",
+        "axes.edgecolor": "#5A5A5A", "axes.linewidth": 0.9,
+        "axes.titlesize": 11.5, "axes.titleweight": "bold",
+        "axes.labelsize": 10.5, "xtick.labelsize": 9,
+        "ytick.labelsize": 9, "legend.fontsize": 9,
+        "font.size": 10, "axes.spines.top": False, "axes.spines.right": False,
+        "grid.color": "#B0B0B0", "grid.alpha": 0.3, "grid.linestyle": ":",
+        "savefig.dpi": 220, "savefig.bbox": "tight",
+    })
+
+
+def _save(fig, name):
+    OUT.mkdir(parents=True, exist_ok=True)
+    for ext in ("png", "pdf"):
+        fig.savefig(OUT / f"{name}.{ext}")
+    plt.close(fig)
+    print("wrote", OUT / f"{name}.png")
+
+
+def _read(path):
+    with path.open() as fh:
+        return list(csv.DictReader(fh))
+
+
+# ---------------------------------------------------------------------------
+def figure_i():
+    rows = _read(ANALYSIS / "deepcmp__per_client.csv")
+    fig, axes = plt.subplots(2, 3, figsize=(14.0, 7.4), constrained_layout=True,
+                             sharey=True)
+    ylim = 0.47
+    for ri, ds in enumerate(DATASETS):
+        for ci, alpha in enumerate(ALPHAS):
+            ax = axes[ri][ci]
+            ax.set_axisbelow(True)
+            ax.grid(axis="y")
+            for seed in (0, 1, 2):
+                sub = [r for r in rows if r["dataset"] == ds
+                       and abs(float(r["alpha"]) - alpha) < 1e-9
+                       and int(r["seed"]) == seed]
+                xs = [int(r["client_id"]) for r in sub]
+                ys = [float(r["delta_rare_f1"]) for r in sub]
+                cols = [TEAL if y > TOL else (RED if y < -TOL else GREY) for y in ys]
+                ax.scatter(xs, ys, s=48, marker=SEED_MARKERS[seed], c=cols,
+                           edgecolor="#333333", linewidth=0.5, alpha=0.9, zorder=4)
+            ax.axhspan(-TOL, TOL, color="#DDDDDD", alpha=0.35, zorder=1)
+            ax.axhline(0.0, color="#555555", lw=1.0, zorder=2)
+            beta = {"xiiotid": {0.25: 0.0, 0.5: 0.75, 0.75: 0.5},
+                    "5gnidd": {0.25: 0.0, 0.5: 1.0, 0.75: 0.5}}[ds][alpha]
+            deltas = [float(r["delta_rare_f1"]) for r in rows
+                      if r["dataset"] == ds and abs(float(r["alpha"]) - alpha) < 1e-9]
+            ax.set_ylim(-ylim, ylim)
+            ax.set_xlim(-0.7, 9.7)
+            ax.set_xticks(range(10))
+            ax.set_title(f"{DATASET_NAME[ds]}  \u03b1={alpha:.2f}   "
+                         f"(beta={beta:g})")
+            ax.text(0.98, 0.04, f"mean \u0394={statistics.mean(deltas):+.3f}",
+                    transform=ax.transAxes, ha="right", va="bottom",
+                    fontsize=8.5, color="#333333")
+            if ri == 1:
+                ax.set_xlabel("client id (per-seed partition)")
+            if ci == 0:
+                ax.set_ylabel("test rare-class macro-F1\n\u0394 (proposed \u2212 original)")
+    handles = [plt.Line2D([0], [0], marker="o", ls="", ms=7, mfc=TEAL, mec="#333", label="improved (> +0.02)"),
+               plt.Line2D([0], [0], marker="o", ls="", ms=7, mfc=RED, mec="#333", label="degraded (< -0.02)"),
+               plt.Line2D([0], [0], marker="o", ls="", ms=7, mfc=GREY, mec="#333", label="unchanged (|\u0394| \u2264 0.02)"),
+               plt.Line2D([0], [0], marker="o", ls="", ms=7, mfc="white", mec="#333", label="seed 0"),
+               plt.Line2D([0], [0], marker="s", ls="", ms=7, mfc="white", mec="#333", label="seed 1"),
+               plt.Line2D([0], [0], marker="^", ls="", ms=7, mfc="white", mec="#333", label="seed 2")]
+    fig.legend(handles=handles, loc="upper center", bbox_to_anchor=(0.5, 1.05),
+               ncols=6, frameon=True, edgecolor="#CCCCCC")
+    fig.suptitle("Paired per-client test rare-class macro-F1: proposed (selected beta) vs ORIGINAL PROTEAN",
+                 fontsize=12.5, weight="bold", y=1.11)
+    fig.text(0.5, -0.03,
+             "Each point is one (client, seed) test observation; x = client id, marker = seed. Points compare the "
+             "support_weighted run at the validation-selected beta against original_protean on the same partition. "
+             "At \u03b1=0.25 the selected beta is 0, so the residual spread is numerical (equal-weight equivalence), not a weighting effect. "
+             "Do not treat client id across seeds as the same real-world client.",
+             ha="center", va="top", fontsize=8.4, color="#444444", linespacing=1.5)
+    _save(fig, "figI_per_client_original_vs_proposed")
+
+
+# ---------------------------------------------------------------------------
+def figure_j():
+    rows = _read(ANALYSIS / "deepcmp__per_class.csv")
+    fig, axes = plt.subplots(2, 2, figsize=(13.4, 8.2), constrained_layout=True)
+    for ri, ds in enumerate(DATASETS):
+        for ci, alpha in enumerate([0.5, 0.75]):
+            ax = axes[ri][ci]
+            ax.set_axisbelow(True)
+            ax.grid(axis="x")
+            stats = defaultdict(lambda: {"f1": [], "sup": []})
+            for r in rows:
+                if r["dataset"] == ds and abs(float(r["alpha"]) - alpha) < 1e-9:
+                    stats[r["class_name"]]["f1"].append(float(r["delta_f1"]))
+                    stats[r["class_name"]]["sup"].append(int(r["support"]))
+            names = sorted(stats, key=lambda n: statistics.mean(stats[n]["f1"]))
+            ys = np.arange(len(names))
+            vals = [statistics.mean(stats[n]["f1"]) for n in names]
+            sup = [int(statistics.mean(stats[n]["sup"])) for n in names]
+            cols = [TEAL if v >= 0 else RED for v in vals]
+            ax.barh(ys, vals, color=cols, edgecolor="#333333", linewidth=0.6, zorder=3)
+            ax.axvline(0.0, color="#555555", lw=1.0, zorder=2)
+            ax.set_yticks(ys)
+            ax.set_yticklabels([f"{n}  (n\u2248{s:,})" for n, s in zip(names, sup)],
+                               fontsize=8.2)
+            beta = {"xiiotid": {0.5: 0.75, 0.75: 0.5},
+                    "5gnidd": {0.5: 1.0, 0.75: 0.5}}[ds][alpha]
+            ax.set_title(f"{DATASET_NAME[ds]}  \u03b1={alpha:.2f}   beta={beta:g}")
+            ax.set_xlabel("\u0394 F1 (proposed \u2212 original), test")
+    fig.suptitle("Per-class test F1 change for beta>0 conditions (mean over seeds {0,1,2})",
+                 fontsize=12.5, weight="bold", y=1.04)
+    fig.text(0.5, -0.015,
+             "Only beta>0 conditions are shown; \u03b1=0.25 selected beta=0 so it carries no weighting effect. "
+             "Class support n\u2248 is the mean test support over seeds. Teal = F1 increased, red = F1 decreased. "
+             "X-IIoTID and 5G-NIDD class labels are never mixed. No significance testing (3 seeds).",
+             ha="center", va="top", fontsize=8.4, color="#444444", linespacing=1.5)
+    _save(fig, "figJ_per_class_original_vs_proposed")
+
+
+# ---------------------------------------------------------------------------
+def figure_k(dataset="5gnidd", alpha=0.75, seed=0, beta=0.5):
+    odir = RESULTS / f"{dataset}__alpha{alpha:g}__seed{seed}__original_protean"
+    sdir = RESULTS / f"{dataset}__alpha{alpha:g}__seed{seed}__support_weighted__beta{beta:g}"
+    names = json.loads((odir / "metadata.json").read_text())["class_names"]
+    o = np.array(json.loads((odir / "test_confusion_matrix.json").read_text()), float)
+    s = np.array(json.loads((sdir / "test_confusion_matrix.json").read_text()), float)
+    diff = s - o
+    orow = np.where(o.sum(1, keepdims=True) == 0, 1, o.sum(1, keepdims=True))
+    srow = np.where(s.sum(1, keepdims=True) == 0, 1, s.sum(1, keepdims=True))
+    diff_rate = s / srow - o / orow
+
+    fig, axes = plt.subplots(1, 2, figsize=(14.0, 6.4), constrained_layout=True)
+    vmax_c = np.abs(diff).max()
+    vmax_r = np.abs(diff_rate).max() * 100
+    im0 = axes[0].imshow(diff, cmap="RdBu_r",
+                         norm=TwoSlopeNorm(vcenter=0, vmin=-vmax_c, vmax=vmax_c))
+    im1 = axes[1].imshow(diff_rate * 100, cmap="RdBu_r",
+                         norm=TwoSlopeNorm(vcenter=0, vmin=-vmax_r, vmax=vmax_r))
+    for ax in axes:
+        ax.set_xticks(range(len(names)))
+        ax.set_yticks(range(len(names)))
+        ax.set_xticklabels(names, rotation=45, ha="right", fontsize=8)
+        ax.set_yticklabels(names, fontsize=8)
+        ax.set_xlabel("predicted class")
+    axes[0].set_ylabel("true class")
+    axes[0].set_title("A. Difference in raw counts\n(proposed \u2212 original)")
+    axes[1].set_title("B. Difference in row-normalized rate\n(percentage points, proposed \u2212 original)")
+    # annotate the strongest cells on each panel
+    for ax, M, is_rate in ((axes[0], diff, False), (axes[1], diff_rate * 100, True)):
+        flat = [(abs(M[i, j]), i, j) for i in range(9) for j in range(9) if i != j]
+        for _, i, j in sorted(flat, reverse=True)[:6]:
+            v = M[i, j]
+            ax.text(j, i, f"{v:+.0f}" if not is_rate else f"{v:+.1f}",
+                    ha="center", va="center", fontsize=6.4, fontweight="bold",
+                    color="white" if abs(v) > (vmax_c if not is_rate else vmax_r) * 0.5
+                    else "#222222")
+    cb0 = fig.colorbar(im0, ax=axes[0], fraction=0.046, pad=0.02)
+    cb0.set_label("count delta")
+    cb1 = fig.colorbar(im1, ax=axes[1], fraction=0.046, pad=0.02)
+    cb1.set_label("row-rate delta (pp)")
+    dname = DATASET_NAME[dataset]
+    fig.suptitle(f"Confusion-matrix change: proposed (beta={beta:g}) minus original PROTEAN - "
+                 f"{dname}, alpha={alpha:g}, seed={seed}, test",
+                 fontsize=12, weight="bold", y=1.05)
+    fig.text(0.5, -0.02,
+             "Rows = true class, columns = predicted class; identical class order and sample counts (243,177) for both runs. "
+             "Diagonal: positive = more correct predictions; off-diagonal: positive = more errors of that (true\u2192predicted) type. "
+             "Red = positive delta (proposed has MORE samples in the cell); blue = negative delta (fewer). Values are descriptive only; no significance is inferred.",
+             ha="center", va="top", fontsize=8.4, color="#444444", linespacing=1.5)
+    _save(fig, f"figK_confusion_delta_{dataset}_a{alpha:g}_seed{seed}")
+
+
+def main():
+    _style()
+    figure_i()
+    figure_j()
+    figure_k()
+
+
+if __name__ == "__main__":
+    main()