TabBench v2: GBDT ensembles, perfmap refresh, dark mode, win-rate view

- GBDT ensembles backfilled from cache (188/188 datasets); single-fit
XGBoost/CatBoost/LightGBM hidden from the leaderboard.
- Performance map: PCA basis frozen to perfmap_freeze.json so re-runs are
stable; co-clusters renamed with 3-part Leader/Runner-up/Composition
template; Small binary cluster rendered as dashed rings + cartouche;
hull smoothing boosts outward push at sharp vertices; pixel-mode axis
arrows that don't break across resolutions.
- Section 04: clusters and trends rendered as two-column rows inside
indented tinted section cards; trend scalars use real per-cluster
ranges with blue/red chip pairs; heatmap bounds tightened.
- Pick-a-model promoted to top-level section 05; depth-3 decision trees
re-fitted on all 188 datasets, post-pruned to drop same-class siblings,
feature and family labels humanised, Bezier connectors with yes/no
badges, cells sized per text length.
- Leaderboard: new Win-rate view (NxN pairwise heatmap by metric,
precomputed alongside bars/table in figures_cache.json).
- Dark mode: removed @media (prefers-color-scheme: dark) overrides so OS
preference never leaks; body.dark is the sole source of truth. Plot
cards lifted; cluster/trend tints bumped; medal cells use dark ink on
brighter coloured bg; tree split-node text and arrow labels picked for
legibility on both themes; handwritten light-mode nudge in dark only.
- Methodology: simplified the CI prose; intro mentions the analysis
tools (leaderboard / win-rate / perfmap / pick-a-model).

base_datasets.json

@@ -26,6 +26,8 @@
   "openml-1487",
   "openml-1489",
   "openml-1491",
+  "openml-1492",
+  "openml-1493",
   "openml-1494",
   "openml-1496",
   "openml-1497",

build_figures_cache.py

@@ -26,6 +26,7 @@ ns = {"__file__": os.path.join(HERE, "app.py"), "__name__": "__main__"}
 exec(compile(prefix, "app.py", "exec"), ns)
 
 plot_global_model_ranking_plotly = ns["plot_global_model_ranking_plotly"]
+plot_winrate_matrix = ns["plot_winrate_matrix"]
 public_per_dataset = ns["public_per_dataset"]
 public_enter_per_dataset = ns["public_enter_per_dataset"]
 compute_elo_for_subset = ns["compute_elo_for_subset"]
@@ -55,27 +56,37 @@ public_per_dataset_with_meta = public_per_dataset.merge(
 )
 
 def build_one(industry, metric):
-    """Return (figure_dict, agg_records, n_datasets)."""
+    """Return (bars_dict, winrate_dict, agg_records, n_datasets)."""
     df = public_per_dataset_with_meta
     if industry != "All":
         df = df[df['dataset_industry'] == industry]
     if df.empty:
-        return None, [], 0
+        return None, None, [], 0
     df = compute_elo_for_subset(df, metric="Accuracy")
     df = compute_pct_improvement_over_baseline(df, baseline_model="xgboost", metric="Accuracy")
     agg = df.groupby('model')[['Accuracy', 'AUC', 'F1_score', 'Precision', 'Recall', 'Cross_entropy', 'Elo_score', '%↗ over XGBoost']].mean().reset_index()
-    fig = plot_global_model_ranking_plotly(agg, metric=metric, per_dataset_df=df)
-    return fig.to_plotly_json(), agg.round(6).to_dict(orient="records"), int(df['dataset_id'].nunique())
+    bars = plot_global_model_ranking_plotly(agg, metric=metric, per_dataset_df=df)
+    # The win-rate matrix only depends on the per-dataset metric and the
+    # filtered subset — no need for Elo / improvement columns.
+    winrate = plot_winrate_matrix(df, metric=metric) if metric not in ("Elo_score", "%↗ over XGBoost") else None
+    return (
+        bars.to_plotly_json(),
+        winrate.to_plotly_json() if winrate is not None else None,
+        agg.round(6).to_dict(orient="records"),
+        int(df['dataset_id'].nunique()),
+    )
 
 cache = {}
 total = len(INDUSTRY_BUCKETS) * len(METRICS)
 i = 0
 for industry in INDUSTRY_BUCKETS:
-    cache[industry] = {"metrics": {}, "n_datasets": 0, "table": []}
+    cache[industry] = {"metrics": {}, "winrate": {}, "n_datasets": 0, "table": []}
     for metric in METRICS:
         i += 1
-        fig_dict, agg, n = build_one(industry, metric)
-        cache[industry]["metrics"][metric] = fig_dict
+        bars_dict, winrate_dict, agg, n = build_one(industry, metric)
+        cache[industry]["metrics"][metric] = bars_dict
+        if winrate_dict is not None:
+            cache[industry]["winrate"][metric] = winrate_dict
         cache[industry]["n_datasets"] = n
         # Save the aggregate table once per industry (same for all metrics)
         if not cache[industry]["table"]:

cluster_overrides.json

@@ -1,22 +1,16 @@
 {
-  "_help": "Map dataset_id (string, without 'openml-' prefix) → cluster name. Restart app to apply.",
-  "_valid_names": [
-    "Foundation-model territory",
-    "Small Industry/Science",
-    "Finance/classical mid-size",
-    "Convergence zone",
-    "LGBM niche",
-    "drop",
-    "blacklist"
-  ],
-  "_help_blacklist": "blacklisted points stay visible as dots but are excluded from EVERY hull. Use either '<id>: \"blacklist\"' or list under '_blacklist': [...]",
-  "_example": {
-    "42636": "Foundation-model territory",
-    "44966": "drop",
-    "12345": "blacklist"
+  "_help": "Map dataset_id (string, without 'openml-' prefix) → cluster rank or name. Restart app to apply.",
+  "_cluster_ranks": {
+    "0": "2nd-gen FM Supremacy / 1st-gen collapses / Mid-size",
+    "1": "2nd-gen FM Supremacy / 1st-gen holds / Small numeric",
+    "2": "GBDT Supremacy / 2nd-gen FMs runner-up / Mid-size binary",
+    "3": "2nd-gen FM Supremacy / Mixed runner-ups / Small binary",
+    "4": "TabPFN v3 lead / Tuned NNs close / Multi-class large"
   },
-  "_help_anchor": "anchored points are forced ONTO the hull as vertices. Format: '_anchor': {'<id>': '<cluster_name>'}",
-  "_anchor": {"45075": "Foundation-model territory"},
-  "_blacklist": ["1444", "23", "1063"],
-  "45075": "Foundation-model territory"
+  "_help_blacklist": "blacklisted points stay visible as dots but are excluded from EVERY hull. List under '_blacklist': [...]",
+  "_help_anchor": "anchored points are forced ONTO the hull as vertices. Format: '_anchor': {'<id>': <rank>}",
+  "_anchor": {
+    "182": 0
+  },
+  "_blacklist": ["46941", "1487", "40981", "46922", "372", "46955", "40678", "741", "23", "1509", "46920", "46937", "1038", "1492"]
 }

perfmap_freeze.json

@@ -0,0 +1 @@
+{"dataset_ids": ["1037", "1038", "1043", "1046", "1049", "1050", "1063", "1067", "1068", "11", "1120", "12", "1220", "137", "14", "1444", "1459", "1462", "1464", "1471", "1475", "1476", "1480", "1481", "1487", "1489", "1491", "1492", "1493", "1494", "1496", "1497", "15", "1501", "1509", "1510", "1548", "1552", "1557", "16", "179", "18", "181", "182", "184", "185", "188", "20", "22", "23", "23381", "251", "255", "28", "29", "3", "30", "301", "31", "310", "32", "334", "335", "36", "37", "372", "40478", "40497", "40499", "40646", "40649", "40664", "40668", "40670", "40677", "40678", "40685", "40701", "40707", "40708", "40900", "40966", "40975", "40981", "40982", "40983", "40994", "41027", "41143", "41156", "41169", "41671", "41705", "41721", "41865", "41875", "41882", "41972", "42178", "42464", "42636", "43039", "43582", "43672", "43812", "43889", "43892", "43895", "44", "44089", "44122", "44123", "44124", "44130", "44150", "44160", "44162", "44186", "44489", "44528", "44966", "44967", "44971", "44972", "45040", "45052", "45062", "45064", "45067", "45075", "4538", "45536", "45537", "45538", "45539", "45540", "45547", "45553", "45558", "45578", "458", "46", "46356", "46441", "46597", "469", "46905", "46906", "46908", "46910", "46911", "46912", "46916", "46919", "46920", "46922", "46924", "46925", "46927", "46929", "46930", "46932", "46933", "46935", "46937", "46938", "46940", "46941", "46944", "46947", "46950", "46951", "46955", "46956", "46958", "46960", "46962", "46963", "46969", "46979", "46980", "54", "6", "60", "6332", "741", "803", "930"], "feature_cols": ["Accuracy/catboost_ensemble", "Accuracy/lightgbm_ensemble", "Accuracy/limix", "Accuracy/mitra", "Accuracy/modern_nca", "Accuracy/realmlp", "Accuracy/seldon", "Accuracy/tabdpt", "Accuracy/tabicl", "Accuracy/tabicl_v2", "Accuracy/tabm", "Accuracy/tabpfn", "Accuracy/tabpfn_v3", "Accuracy/xgboost_ensemble", "AUC/catboost_ensemble", "AUC/lightgbm_ensemble", "AUC/limix", "AUC/mitra", "AUC/modern_nca", "AUC/realmlp", "AUC/seldon", "AUC/tabdpt", "AUC/tabicl", "AUC/tabicl_v2", "AUC/tabm", "AUC/tabpfn", "AUC/tabpfn_v3", "AUC/xgboost_ensemble", "F1_score/catboost_ensemble", "F1_score/lightgbm_ensemble", "F1_score/limix", "F1_score/mitra", "F1_score/modern_nca", "F1_score/realmlp", "F1_score/seldon", "F1_score/tabdpt", "F1_score/tabicl", "F1_score/tabicl_v2", "F1_score/tabm", "F1_score/tabpfn", "F1_score/tabpfn_v3", "F1_score/xgboost_ensemble", "Precision/catboost_ensemble", "Precision/lightgbm_ensemble", "Precision/limix", "Precision/mitra", "Precision/modern_nca", "Precision/realmlp", "Precision/seldon", "Precision/tabdpt", "Precision/tabicl", "Precision/tabicl_v2", "Precision/tabm", "Precision/tabpfn", "Precision/tabpfn_v3", "Precision/xgboost_ensemble", "Recall/catboost_ensemble", "Recall/lightgbm_ensemble", "Recall/limix", "Recall/mitra", "Recall/modern_nca", "Recall/realmlp", "Recall/seldon", "Recall/tabdpt", "Recall/tabicl", "Recall/tabicl_v2", "Recall/tabm", "Recall/tabpfn", "Recall/tabpfn_v3", "Recall/xgboost_ensemble", "Cross_entropy/catboost_ensemble", "Cross_entropy/lightgbm_ensemble", "Cross_entropy/limix", "Cross_entropy/mitra", "Cross_entropy/modern_nca", "Cross_entropy/realmlp", "Cross_entropy/seldon", "Cross_entropy/tabdpt", "Cross_entropy/tabicl", "Cross_entropy/tabicl_v2", "Cross_entropy/tabm", "Cross_entropy/tabpfn", "Cross_entropy/tabpfn_v3", "Cross_entropy/xgboost_ensemble"], "feature_mean": [0.4464745656235018, 0.3296590082494337, 0.633882296781233, 0.5831117021276596, 0.2686537014994462, 0.26101667245284266, 0.7412471689067434, 0.2539565044884194, 0.6104112937889534, 0.7587490080841144, 0.36121805204252, 0.6199510448180662, 0.7362367537899454, 0.39543222734712097, 0.4683432111623601, 0.3373127758766057, 0.6410085924713584, 0.5817478177850518, 0.27595099935525463, 0.1909429813685133, 0.7803746342310172, 0.2110279348311263, 0.6131177776124584, 0.7993241539784093, 0.3218432276942915, 0.6151813759526525, 0.75916354626461, 0.40466097141629054, 0.4571223693564119, 0.3377743267370927, 0.6209486435550267, 0.5377966448445173, 0.35379242506902087, 0.3477553770106962, 0.7125660244011308, 0.25672529302848446, 0.5644291730727901, 0.7690224668948072, 0.38224480897353236, 0.5882999388318537, 0.6928042701978874, 0.37871823802674875, 0.46008530476615583, 0.3376296723371191, 0.6222892798690671, 0.5815091380250954, 0.30597804972804976, 0.3214299294086528, 0.6935071872571872, 0.2984158276711468, 0.5864820463224718, 0.7064357775793946, 0.38357934913785974, 0.5883498445998446, 0.6826051017540379, 0.4317034915439171, 0.4419783968986097, 0.32605380730380734, 0.6118555646481177, 0.5199979541734862, 0.38952857131048624, 0.3737351005968028, 0.7036391947296203, 0.3083442857379028, 0.564123745639703, 0.7273434012795714, 0.4119019573806807, 0.5775568698441038, 0.6751923903519648, 0.36874876010514307, 0.4331628477905073, 0.36182735704012303, 0.6380591925804692, 0.6103041462084016, 0.2736121443568252, 0.12750438096182776, 0.8045562003008813, 0.21179697961612853, 0.6335213840532989, 0.8149620178875498, 0.3055555555555555, 0.6325978690340393, 0.7684789796491923, 0.3840609449652003], "pca_components": [[0.146907808398491, 0.20114869932291057, -0.052711099432679764, 0.0051444468351336675, -0.01778875692716969, -0.009608283538971639, -0.12726378240610162, -0.10120227345499802, -0.07887575573082464, -0.13611213758361349, 0.13388309847107743, -0.033969240711765415, -0.11375517300681823, 0.18420244976532957, 0.13722421644834, 0.17842629313836803, -0.051431945521123686, 0.00421693608698639, -0.011449385020097716, 0.004066285346033057, -0.11229061555184228, -0.10140800618074952, -0.0625721680059459, -0.12631434932597596, 0.09433547339142263, -0.037490867409539955, -0.08621377682879586, 0.1709019094329207, 0.14262563556209817, 0.20054956628004442, -0.06864127827615504, -0.011480807229687426, 0.02832214122100685, 0.04880339903172603, -0.13999319320477277, -0.07998978759623836, -0.10848346733658146, -0.11119772904080946, 0.1387118417070701, -0.06686126565295093, -0.14804828476914933, 0.17568322930439903, 0.13823388568345177, 0.18715388340658645, -0.04324099256732326, 0.000570773475366898, 0.024999058359134917, 0.029175299284920058, -0.13418199700359057, -0.07262325846948535, -0.0899423540213617, -0.15471303984684923, 0.12035787889742627, -0.044942780600894176, -0.13496681099914198, 0.17412045440176002, 0.13128067417200553, 0.19111324923149253, -0.06298802581546631, -0.0080849760791246, 0.032387163871303616, 0.042473631813694804, -0.1402985265359515, -0.07996941303500239, -0.09652823428896147, -0.1326428713724256, 0.14848769205702228, -0.05108459695144007, -0.1439318953579276, 0.16978612829078077, 0.11754251135232524, 0.15893983748618834, -0.04723326009306483, 0.02465284381169231, -0.024622301322837863, 0.00917230244314059, -0.09852064596630367, -0.07270387115643288, -0.06420928488424045, -0.11722993483774771, 0.0937290243794225, -0.030146333442777156, -0.09110081042295107, 0.14172992265358675], [-0.08491764389561016, -0.11266717097808268, -0.12221766679316537, 0.02515544726370378, 0.10833740913104484, 0.26721139102096275, -0.01692317651774755, -0.10728204493639636, -0.08809135811439489, -0.005432587268653479, 0.21096327665713197, -0.05481016776946194, 0.0699823308382295, -0.0893080386375602, -0.07635120372854177, -0.09753048362601836, -0.11486706071507286, 0.005061944145388893, 0.11157828762143542, 0.23820399786604007, -0.016071396241339016, -0.04641202670659471, -0.07294032913587536, -0.011692676455840233, 0.1872756325979318, -0.05595102180755615, 0.05665463578158662, -0.10695829959554437, -0.10629494799090607, -0.11374216767513866, -0.11386834137575629, 0.03260558564748944, 0.11575162494523585, 0.27431708496190566, -0.018918520681540367, -0.09569582248675142, -0.08607273058721597, 0.00632032996594563, 0.1889601322113517, -0.06252067676562491, 0.06698233848664495, -0.08782388865563959, -0.0754063819998581, -0.10667675368067, -0.11932076495445153, 0.009889751984172985, 0.10515455988024482, 0.24390039605549513, 0.004803013049359491, -0.13081696261636516, -0.08677528788994615, 0.008420821362319466, 0.19475096625513633, -0.053571881713062555, 0.07480750402389072, -0.06915897975626556, -0.07352638915831662, -0.11013654823850225, -0.07278844862745802, 0.03503264937020091, 0.08180161768517111, 0.24005619844212742, -0.006255216713246687, -0.11606099899424899, -0.07548552034330429, -0.0008236341818161691, 0.1724061287219457, -0.0625879157782033, 0.057373332959452523, -0.0690052551438013, -0.04405994155289653, -0.11118059744004193, -0.11742867748730677, 0.011028806201351531, 0.11289256668410291, 0.12732648140393302, -0.007148125273953241, -0.055869863276458834, -0.059778308663241415, 0.0049810717624958626, 0.2066469238738026, -0.05433801974170208, 0.06566617826684029, -0.07873849475692536]], "pca_mean": [0.4464745656235018, 0.3296590082494337, 0.633882296781233, 0.5831117021276596, 0.2686537014994462, 0.26101667245284266, 0.7412471689067434, 0.2539565044884194, 0.6104112937889534, 0.7587490080841144, 0.36121805204252, 0.6199510448180662, 0.7362367537899454, 0.39543222734712097, 0.4683432111623601, 0.3373127758766057, 0.6410085924713584, 0.5817478177850518, 0.27595099935525463, 0.1909429813685133, 0.7803746342310172, 0.2110279348311263, 0.6131177776124584, 0.7993241539784093, 0.3218432276942915, 0.6151813759526525, 0.75916354626461, 0.40466097141629054, 0.4571223693564119, 0.3377743267370927, 0.6209486435550267, 0.5377966448445173, 0.35379242506902087, 0.3477553770106962, 0.7125660244011308, 0.25672529302848446, 0.5644291730727901, 0.7690224668948072, 0.38224480897353236, 0.5882999388318537, 0.6928042701978874, 0.37871823802674875, 0.46008530476615583, 0.3376296723371191, 0.6222892798690671, 0.5815091380250954, 0.30597804972804976, 0.3214299294086528, 0.6935071872571872, 0.2984158276711468, 0.5864820463224718, 0.7064357775793946, 0.38357934913785974, 0.5883498445998446, 0.6826051017540379, 0.4317034915439171, 0.4419783968986097, 0.32605380730380734, 0.6118555646481177, 0.5199979541734862, 0.38952857131048624, 0.3737351005968028, 0.7036391947296203, 0.3083442857379028, 0.564123745639703, 0.7273434012795714, 0.4119019573806807, 0.5775568698441038, 0.6751923903519648, 0.36874876010514307, 0.4331628477905073, 0.36182735704012303, 0.6380591925804692, 0.6103041462084016, 0.2736121443568252, 0.12750438096182776, 0.8045562003008813, 0.21179697961612853, 0.6335213840532989, 0.8149620178875498, 0.3055555555555555, 0.6325978690340393, 0.7684789796491923, 0.3840609449652003], "row_labels": [0, 3, 0, 0, 1, 4, 0, 1, 1, 2, 3, 3, 0, 0, 3, 1, 1, 4, 4, 1, 1, 1, 4, 2, 1, 1, 3, 2, 2, 1, 3, 0, 3, 3, 3, 1, 3, 3, 3, 1, 0, 3, 1, 3, 2, 1, 1, 3, 1, 2, 1, 0, 0, 3, 0, 3, 1, 4, 3, 1, 3, 2, 4, 3, 0, 0, 3, 0, 2, 3, 3, 2, 2, 3, 0, 0, 3, 2, 3, 3, 1, 1, 2, 1, 3, 3, 0, 2, 3, 0, 2, 3, 3, 1, 4, 1, 1, 3, 3, 1, 3, 0, 3, 1, 1, 0, 0, 0, 3, 3, 3, 3, 3, 0, 0, 3, 4, 0, 4, 4, 4, 1, 1, 1, 3, 3, 3, 4, 0, 0, 1, 1, 1, 1, 1, 1, 0, 3, 1, 3, 1, 3, 3, 2, 3, 0, 4, 1, 0, 0, 3, 0, 4, 0, 3, 0, 4, 4, 4, 0, 1, 3, 0, 0, 3, 4, 1, 1, 2, 3, 3, 1, 0, 4, 3, 3, 0, 1, 1, 4, 0, 1, 2, 3, 0, 2, 3, 1], "col_labels": [0, 0, 3, 3, 4, 2, 3, 4, 1, 3, 2, 3, 3, 0, 0, 0, 3, 3, 4, 2, 3, 4, 1, 3, 2, 3, 3, 0, 0, 0, 1, 3, 4, 2, 3, 4, 1, 3, 2, 3, 3, 0, 0, 0, 3, 3, 4, 2, 3, 4, 1, 3, 2, 3, 3, 0, 0, 0, 1, 3, 4, 2, 3, 4, 1, 3, 2, 3, 3, 0, 0, 0, 3, 3, 4, 2, 3, 4, 1, 3, 2, 3, 3, 0]}

public_datasets_info.json

@@ -3380,7 +3380,12 @@
   {
     "dataset_id": "43889",
     "n_classes": 2,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "law-school-admission-bianry",
+    "features": 11,
+    "rows": 20800,
+    "pct_cat": 0.364,
+    "avg_card": 2.8
   },
   {
     "dataset_id": "151",
@@ -3395,12 +3400,22 @@
   {
     "dataset_id": "41865",
     "n_classes": 2,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "FOREX_audsgd-hour-High",
+    "features": 11,
+    "rows": 43825,
+    "pct_cat": 0.0,
+    "avg_card": 0
   },
   {
     "dataset_id": "41875",
     "n_classes": 2,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "FOREX_audchf-day-High",
+    "features": 11,
+    "rows": 1833,
+    "pct_cat": 0.0,
+    "avg_card": 0
   },
   {
     "dataset_id": "41839",
@@ -3415,7 +3430,12 @@
   {
     "dataset_id": "44967",
     "n_classes": 17,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "student_performance_por",
+    "features": 30,
+    "rows": 649,
+    "pct_cat": 0.533,
+    "avg_card": 2.53
   },
   {
     "dataset_id": "41718",
@@ -3430,17 +3450,32 @@
   {
     "dataset_id": "54",
     "n_classes": 4,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "vehicle",
+    "features": 18,
+    "rows": 846,
+    "pct_cat": 0.0,
+    "avg_card": 0
   },
   {
     "dataset_id": "469",
     "n_classes": 6,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "analcatdata_dmft",
+    "features": 4,
+    "rows": 797,
+    "pct_cat": 1.0,
+    "avg_card": 5.25
   },
   {
     "dataset_id": "188",
     "n_classes": 5,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "eucalyptus",
+    "features": 19,
+    "rows": 736,
+    "pct_cat": 0.263,
+    "avg_card": 15.4
   },
   {
     "dataset_id": "45068",
@@ -3450,12 +3485,22 @@
   {
     "dataset_id": "44966",
     "n_classes": 8,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "solar_flare",
+    "features": 10,
+    "rows": 1066,
+    "pct_cat": 0.7,
+    "avg_card": 3.38
   },
   {
     "dataset_id": "458",
     "n_classes": 4,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "analcatdata_authorship",
+    "features": 70,
+    "rows": 841,
+    "pct_cat": 0.0,
+    "avg_card": 0
   },
   {
     "dataset_id": "1044",
@@ -3465,7 +3510,12 @@
   {
     "dataset_id": "41882",
     "n_classes": 2,
-    "source": "openml_only"
+    "source": "openml_only",
+    "name": "FOREX_audjpy-day-High",
+    "features": 11,
+    "rows": 1832,
+    "pct_cat": 0.0,
+    "avg_card": 0
   },
   {
     "dataset_id": "41763",

results_academic_new/catboost.json

@@ -2003,6 +2003,133 @@
           "mem_usage": 7263.78125,
           "time_usage": 0.4109564699992916,
           "fold": 0,
+          "imputed": false,
+          "test_ensemble_cross_entropy": 0.43386434873386975,
+          "test_ensemble_roc_auc": 0.8755555555555555,
+          "test_ensemble_precision": 0.7692307692307693,
+          "test_ensemble_recall": 0.5555555555555556,
+          "test_ensemble_hinge_loss": 0.953825658629945,
+          "test_ensemble_f1_score": 0.6451612903225806,
+          "test_ensemble_accuracy": 0.7857142857142857
+        },
+        {
+          "test_ensemble_cross_entropy": 0.49149430206183053,
+          "test_ensemble_roc_auc": 0.8261111111111111,
+          "test_ensemble_precision": 0.6666666666666666,
+          "test_ensemble_recall": 0.4444444444444444,
+          "test_ensemble_hinge_loss": 0.9795520750090578,
+          "test_ensemble_f1_score": 0.5333333333333333,
+          "test_ensemble_accuracy": 0.7272727272727273,
+          "test_cross_entropy": 0.49813900797993527,
+          "test_roc_auc": 0.8242592592592592,
+          "test_precision": 0.6829268292682927,
+          "test_recall": 0.5185185185185185,
+          "test_hinge_loss": 0.9537525810288828,
+          "test_f1_score": 0.5894736842105263,
+          "test_accuracy": 0.7467532467532467,
+          "best_hyperopt_params": {
+            "iterations": 1000,
+            "depth": 5,
+            "learning_rate": 0.010728780287812144,
+            "l2_leaf_reg": 8.87502970852228,
+            "random_strength": 8.690112656110974,
+            "bagging_temperature": 0.7609494759223234
+          },
+          "best_hyperopt_score": 0.8355232558139536,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.33092466701054946,
+          "fold": 1,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.43827509997296094,
+          "test_ensemble_roc_auc": 0.8625925925925926,
+          "test_ensemble_precision": 0.74,
+          "test_ensemble_recall": 0.6851851851851852,
+          "test_ensemble_hinge_loss": 0.9384958763550146,
+          "test_ensemble_f1_score": 0.7115384615384616,
+          "test_ensemble_accuracy": 0.8051948051948052,
+          "test_cross_entropy": 0.4264530086053719,
+          "test_roc_auc": 0.8772222222222222,
+          "test_precision": 0.7450980392156863,
+          "test_recall": 0.7037037037037037,
+          "test_hinge_loss": 0.9366071680728787,
+          "test_f1_score": 0.7238095238095238,
+          "test_accuracy": 0.8116883116883117,
+          "best_hyperopt_params": {
+            "iterations": 100,
+            "depth": 4,
+            "learning_rate": 0.2045030922941104,
+            "l2_leaf_reg": 9.08548764129536,
+            "random_strength": 7.291044809941281,
+            "bagging_temperature": 0.3909618858216055
+          },
+          "best_hyperopt_score": 0.8363953488372091,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.01828358302009292,
+          "fold": 2,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.4826366082652167,
+          "test_ensemble_roc_auc": 0.8366666666666667,
+          "test_ensemble_precision": 0.6976744186046512,
+          "test_ensemble_recall": 0.5555555555555556,
+          "test_ensemble_hinge_loss": 0.9724580474039151,
+          "test_ensemble_f1_score": 0.6185567010309279,
+          "test_ensemble_accuracy": 0.7597402597402597,
+          "test_cross_entropy": 0.49094211859795456,
+          "test_roc_auc": 0.8292592592592593,
+          "test_precision": 0.6976744186046512,
+          "test_recall": 0.5555555555555556,
+          "test_hinge_loss": 0.9685461476790421,
+          "test_f1_score": 0.6185567010309279,
+          "test_accuracy": 0.7597402597402597,
+          "best_hyperopt_params": {
+            "iterations": 500,
+            "depth": 3,
+            "learning_rate": 0.019221228697919146,
+            "l2_leaf_reg": 4.257444873996925,
+            "random_strength": 8.540005315561082,
+            "bagging_temperature": 0.10694931956210899
+          },
+          "best_hyperopt_score": 0.8644186046511628,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.11811866701464169,
+          "fold": 3,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.4615906081605936,
+          "test_ensemble_roc_auc": 0.8535185185185185,
+          "test_ensemble_precision": 0.775,
+          "test_ensemble_recall": 0.5740740740740741,
+          "test_ensemble_hinge_loss": 0.9713839236725454,
+          "test_ensemble_f1_score": 0.6595744680851063,
+          "test_ensemble_accuracy": 0.7922077922077922,
+          "test_cross_entropy": 0.4602084387204193,
+          "test_roc_auc": 0.8462962962962963,
+          "test_precision": 0.7380952380952381,
+          "test_recall": 0.5740740740740741,
+          "test_hinge_loss": 0.9511177680256443,
+          "test_f1_score": 0.6458333333333334,
+          "test_accuracy": 0.7792207792207793,
+          "best_hyperopt_params": {
+            "iterations": 100,
+            "depth": 7,
+            "learning_rate": 0.12065395983547453,
+            "l2_leaf_reg": 4.0466198947725935,
+            "random_strength": 5.3490273256053795,
+            "bagging_temperature": 0.6743313880963466
+          },
+          "best_hyperopt_score": 0.801627906976744,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.04541929200058803,
+          "fold": 4,
           "imputed": false
         }
       ],
@@ -2016,7 +2143,14 @@
       "mean_cross_entropy": 0.44017635609996436,
       "mean_precision": 0.7631578947368421,
       "n_folds_imputed": 0,
-      "fully_imputed": false
+      "fully_imputed": false,
+      "mean_ensemble_accuracy": 0.7740259740259741,
+      "mean_ensemble_roc_auc": 0.8508888888888888,
+      "mean_ensemble_f1_score": 0.6336328508620819,
+      "mean_ensemble_precision": 0.7297143709004175,
+      "mean_ensemble_recall": 0.562962962962963,
+      "mean_ensemble_cross_entropy": 0.4615721934388943,
+      "mean_ensemble_hinge_loss": 0.9631431162140955
     },
     "openml-44": {
       "folds": [
@@ -2275,6 +2409,133 @@
           "mem_usage": 8197.94921875,
           "time_usage": 1.282376753999415,
           "fold": 0,
+          "imputed": false,
+          "test_ensemble_cross_entropy": 0.4995297756405721,
+          "test_ensemble_roc_auc": 0.9359551275744457,
+          "test_ensemble_precision": 0.7823529411764706,
+          "test_ensemble_recall": 0.7823529411764706,
+          "test_ensemble_hinge_loss": 0.565280936736336,
+          "test_ensemble_f1_score": 0.7807601609099737,
+          "test_ensemble_accuracy": 0.7823529411764706
+        },
+        {
+          "test_ensemble_cross_entropy": 0.5191902786797484,
+          "test_ensemble_roc_auc": 0.9375084038436311,
+          "test_ensemble_precision": 0.7764705882352941,
+          "test_ensemble_recall": 0.7764705882352941,
+          "test_ensemble_hinge_loss": 0.5966436222840387,
+          "test_ensemble_f1_score": 0.7694984858760627,
+          "test_ensemble_accuracy": 0.7764705882352941,
+          "test_cross_entropy": 0.5592659988573994,
+          "test_roc_auc": 0.9310962711530893,
+          "test_precision": 0.7588235294117647,
+          "test_recall": 0.7588235294117647,
+          "test_hinge_loss": 0.6345513152534152,
+          "test_f1_score": 0.7524346821354462,
+          "test_accuracy": 0.7588235294117647,
+          "best_hyperopt_params": {
+            "iterations": 100,
+            "depth": 4,
+            "learning_rate": 0.1796210586557946,
+            "l2_leaf_reg": 7.989595989147723,
+            "random_strength": 9.987623305612088,
+            "bagging_temperature": 0.09882291645416383
+          },
+          "best_hyperopt_score": 0.924765781312525,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.03567095802281983,
+          "fold": 1,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.409383497505377,
+          "test_ensemble_roc_auc": 0.9499266200970746,
+          "test_ensemble_precision": 0.8176470588235294,
+          "test_ensemble_recall": 0.8176470588235294,
+          "test_ensemble_hinge_loss": 0.4543626718834811,
+          "test_ensemble_f1_score": 0.8162304098549079,
+          "test_ensemble_accuracy": 0.8176470588235294,
+          "test_cross_entropy": 0.4252720235327064,
+          "test_roc_auc": 0.9502972091040274,
+          "test_precision": 0.8058823529411765,
+          "test_recall": 0.8058823529411765,
+          "test_hinge_loss": 0.42110784200675366,
+          "test_f1_score": 0.8041672375993423,
+          "test_accuracy": 0.8058823529411765,
+          "best_hyperopt_params": {
+            "iterations": 1000,
+            "depth": 4,
+            "learning_rate": 0.10750131318142611,
+            "l2_leaf_reg": 1.0726852622519663,
+            "random_strength": 6.99673374646194,
+            "bagging_temperature": 0.2573306107809358
+          },
+          "best_hyperopt_score": 0.9374546443577432,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.34088804098428227,
+          "fold": 2,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.44409359095993534,
+          "test_ensemble_roc_auc": 0.9567265594254231,
+          "test_ensemble_precision": 0.788235294117647,
+          "test_ensemble_recall": 0.788235294117647,
+          "test_ensemble_hinge_loss": 0.5384316865080074,
+          "test_ensemble_f1_score": 0.7877816705766143,
+          "test_ensemble_accuracy": 0.788235294117647,
+          "test_cross_entropy": 0.3920124464331481,
+          "test_roc_auc": 0.9556904269972453,
+          "test_precision": 0.8294117647058824,
+          "test_recall": 0.8294117647058824,
+          "test_hinge_loss": 0.4587006207049688,
+          "test_f1_score": 0.8309535661797075,
+          "test_accuracy": 0.8294117647058824,
+          "best_hyperopt_params": {
+            "iterations": 500,
+            "depth": 3,
+            "learning_rate": 0.1998321787205074,
+            "l2_leaf_reg": 5.761438018916822,
+            "random_strength": 4.195256682086876,
+            "bagging_temperature": 0.9632620429095796
+          },
+          "best_hyperopt_score": 0.9325891106442576,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.13038674998097122,
+          "fold": 3,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.44384802675578183,
+          "test_ensemble_roc_auc": 0.941392619375574,
+          "test_ensemble_precision": 0.8,
+          "test_ensemble_recall": 0.8,
+          "test_ensemble_hinge_loss": 0.4813423391723857,
+          "test_ensemble_f1_score": 0.7989910779126466,
+          "test_ensemble_accuracy": 0.8,
+          "test_cross_entropy": 0.46633499466802614,
+          "test_roc_auc": 0.9357257641348551,
+          "test_precision": 0.7647058823529411,
+          "test_recall": 0.7647058823529411,
+          "test_hinge_loss": 0.49749400287921786,
+          "test_f1_score": 0.7610741561961074,
+          "test_accuracy": 0.7647058823529411,
+          "best_hyperopt_params": {
+            "iterations": 1000,
+            "depth": 4,
+            "learning_rate": 0.12457162713174764,
+            "l2_leaf_reg": 4.813543710238976,
+            "random_strength": 5.1228727156460705,
+            "bagging_temperature": 0.765191625286142
+          },
+          "best_hyperopt_score": 0.9376152961184474,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.33613587499712594,
+          "fold": 4,
           "imputed": false
         }
       ],
@@ -2288,7 +2549,14 @@
       "mean_cross_entropy": 0.47274286765377455,
       "mean_precision": 0.7647058823529411,
       "n_folds_imputed": 0,
-      "fully_imputed": false
+      "fully_imputed": false,
+      "mean_ensemble_accuracy": 0.7929411764705883,
+      "mean_ensemble_roc_auc": 0.9443018660632297,
+      "mean_ensemble_f1_score": 0.790652361026041,
+      "mean_ensemble_precision": 0.7929411764705883,
+      "mean_ensemble_recall": 0.7929411764705883,
+      "mean_ensemble_cross_entropy": 0.46320903390828294,
+      "mean_ensemble_hinge_loss": 0.5272122513168498
     },
     "openml-60": {
       "folds": [
@@ -3279,6 +3547,133 @@
           "mem_usage": 7507.546875,
           "time_usage": 0.5897211499996047,
           "fold": 0,
+          "imputed": false,
+          "test_ensemble_cross_entropy": 0.6769080429035581,
+          "test_ensemble_roc_auc": 0.9307068254742674,
+          "test_ensemble_precision": 0.7364864864864865,
+          "test_ensemble_recall": 0.7364864864864865,
+          "test_ensemble_hinge_loss": 0.7137795841793233,
+          "test_ensemble_f1_score": 0.7145048203190314,
+          "test_ensemble_accuracy": 0.7364864864864865
+        },
+        {
+          "test_ensemble_cross_entropy": 0.6099132956438846,
+          "test_ensemble_roc_auc": 0.9411204697541906,
+          "test_ensemble_precision": 0.7297297297297297,
+          "test_ensemble_recall": 0.7297297297297297,
+          "test_ensemble_hinge_loss": 0.6861820938491757,
+          "test_ensemble_f1_score": 0.7166959386706823,
+          "test_ensemble_accuracy": 0.7297297297297297,
+          "test_cross_entropy": 0.6296871377408355,
+          "test_roc_auc": 0.9356988887512143,
+          "test_precision": 0.6959459459459459,
+          "test_recall": 0.6959459459459459,
+          "test_hinge_loss": 0.6917632292826239,
+          "test_f1_score": 0.6658620895166577,
+          "test_accuracy": 0.6959459459459459,
+          "best_hyperopt_params": {
+            "iterations": 1000,
+            "depth": 3,
+            "learning_rate": 0.038695762894964444,
+            "l2_leaf_reg": 1.0282615984445402,
+            "random_strength": 6.105605356536472,
+            "bagging_temperature": 0.5613187626086535
+          },
+          "best_hyperopt_score": 0.9154723268617792,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 1.265960749995429,
+          "fold": 1,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.800842198637179,
+          "test_ensemble_roc_auc": 0.894479632124981,
+          "test_ensemble_precision": 0.6418918918918919,
+          "test_ensemble_recall": 0.6418918918918919,
+          "test_ensemble_hinge_loss": 0.7993121126415409,
+          "test_ensemble_f1_score": 0.6020976748908519,
+          "test_ensemble_accuracy": 0.6418918918918919,
+          "test_cross_entropy": 0.8242927350653093,
+          "test_roc_auc": 0.8920131978271515,
+          "test_precision": 0.6351351351351351,
+          "test_recall": 0.6351351351351351,
+          "test_hinge_loss": 0.7960461922246324,
+          "test_f1_score": 0.5794183082662345,
+          "test_accuracy": 0.6351351351351351,
+          "best_hyperopt_params": {
+            "iterations": 500,
+            "depth": 7,
+            "learning_rate": 0.0663661833392611,
+            "l2_leaf_reg": 8.914973110896346,
+            "random_strength": 9.046324053570908,
+            "bagging_temperature": 0.15860617248776387
+          },
+          "best_hyperopt_score": 0.9248510235737297,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 1.4197912079980597,
+          "fold": 2,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.6778213556782435,
+          "test_ensemble_roc_auc": 0.9381415096531376,
+          "test_ensemble_precision": 0.7027027027027027,
+          "test_ensemble_recall": 0.7027027027027027,
+          "test_ensemble_hinge_loss": 0.7311685950661799,
+          "test_ensemble_f1_score": 0.6664237640434657,
+          "test_ensemble_accuracy": 0.7027027027027027,
+          "test_cross_entropy": 0.618510961015219,
+          "test_roc_auc": 0.9409317695654906,
+          "test_precision": 0.7297297297297297,
+          "test_recall": 0.7297297297297297,
+          "test_hinge_loss": 0.6777513341562841,
+          "test_f1_score": 0.7098039215686274,
+          "test_accuracy": 0.7297297297297297,
+          "best_hyperopt_params": {
+            "iterations": 1000,
+            "depth": 5,
+            "learning_rate": 0.044773772104018644,
+            "l2_leaf_reg": 4.919129522443123,
+            "random_strength": 6.897168300565475,
+            "bagging_temperature": 0.7268740929392631
+          },
+          "best_hyperopt_score": 0.9224308100522156,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 2.952929041988682,
+          "fold": 3,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.6770786118765768,
+          "test_ensemble_roc_auc": 0.9341897282304259,
+          "test_ensemble_precision": 0.6959459459459459,
+          "test_ensemble_recall": 0.6959459459459459,
+          "test_ensemble_hinge_loss": 0.7184555114540551,
+          "test_ensemble_f1_score": 0.6546247150022799,
+          "test_ensemble_accuracy": 0.6959459459459459,
+          "test_cross_entropy": 0.70340294472938,
+          "test_roc_auc": 0.9264057486441206,
+          "test_precision": 0.6891891891891891,
+          "test_recall": 0.6891891891891891,
+          "test_hinge_loss": 0.70102297338353,
+          "test_f1_score": 0.6689418242213272,
+          "test_accuracy": 0.6891891891891891,
+          "best_hyperopt_params": {
+            "iterations": 1000,
+            "depth": 6,
+            "learning_rate": 0.13429290984610845,
+            "l2_leaf_reg": 7.075510344785677,
+            "random_strength": 3.8009092338708195,
+            "bagging_temperature": 0.2992320448264406
+          },
+          "best_hyperopt_score": 0.910102499417617,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 4.694548083993141,
+          "fold": 4,
           "imputed": false
         }
       ],
@@ -3292,7 +3687,14 @@
       "mean_cross_entropy": 0.666305730311089,
       "mean_precision": 0.6959459459459459,
       "n_folds_imputed": 0,
-      "fully_imputed": false
+      "fully_imputed": false,
+      "mean_ensemble_accuracy": 0.7013513513513514,
+      "mean_ensemble_roc_auc": 0.9277276330474006,
+      "mean_ensemble_f1_score": 0.6708693825852622,
+      "mean_ensemble_precision": 0.7013513513513514,
+      "mean_ensemble_recall": 0.7013513513513514,
+      "mean_ensemble_cross_entropy": 0.6885127009478884,
+      "mean_ensemble_hinge_loss": 0.7297795794380549
     },
     "openml-251": {
       "folds": [
@@ -4161,6 +4563,133 @@
           "mem_usage": 8250.0625,
           "time_usage": 1.080307604999689,
           "fold": 0,
+          "imputed": false,
+          "test_ensemble_cross_entropy": 0.2405747974520093,
+          "test_ensemble_roc_auc": 0.9974865750421896,
+          "test_ensemble_precision": 0.9822485207100592,
+          "test_ensemble_recall": 0.9822485207100592,
+          "test_ensemble_hinge_loss": 0.30736802743258257,
+          "test_ensemble_f1_score": 0.976388607707694,
+          "test_ensemble_accuracy": 0.9822485207100592
+        },
+        {
+          "test_ensemble_cross_entropy": 0.008271910575698342,
+          "test_ensemble_roc_auc": 1.0,
+          "test_ensemble_precision": 1.0,
+          "test_ensemble_recall": 1.0,
+          "test_ensemble_hinge_loss": 0.012890688458434125,
+          "test_ensemble_f1_score": 1.0,
+          "test_ensemble_accuracy": 1.0,
+          "test_cross_entropy": 0.008718858636653415,
+          "test_roc_auc": 1.0,
+          "test_precision": 1.0,
+          "test_recall": 1.0,
+          "test_hinge_loss": 0.014949338548392307,
+          "test_f1_score": 1.0,
+          "test_accuracy": 1.0,
+          "best_hyperopt_params": {
+            "iterations": 1000,
+            "depth": 5,
+            "learning_rate": 0.29983733655034384,
+            "l2_leaf_reg": 1.255911879694784,
+            "random_strength": 9.90045014085742,
+            "bagging_temperature": 0.01347678954496262
+          },
+          "best_hyperopt_score": 1.0,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.5720001249865163,
+          "fold": 1,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.09633398161035646,
+          "test_ensemble_roc_auc": 1.0,
+          "test_ensemble_precision": 1.0,
+          "test_ensemble_recall": 1.0,
+          "test_ensemble_hinge_loss": 0.13904260652524622,
+          "test_ensemble_f1_score": 1.0,
+          "test_ensemble_accuracy": 1.0,
+          "test_cross_entropy": 0.07829736893657216,
+          "test_roc_auc": 1.0,
+          "test_precision": 1.0,
+          "test_recall": 1.0,
+          "test_hinge_loss": 0.11476935113821074,
+          "test_f1_score": 1.0,
+          "test_accuracy": 1.0,
+          "best_hyperopt_params": {
+            "iterations": 100,
+            "depth": 8,
+            "learning_rate": 0.07484596578015762,
+            "l2_leaf_reg": 1.015826702116557,
+            "random_strength": 5.335840456298519,
+            "bagging_temperature": 0.004701791922735399
+          },
+          "best_hyperopt_score": 0.9998364357563355,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.193477291992167,
+          "fold": 2,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.04168186409058239,
+          "test_ensemble_roc_auc": 0.9996404151992077,
+          "test_ensemble_precision": 0.9822485207100592,
+          "test_ensemble_recall": 0.9822485207100592,
+          "test_ensemble_hinge_loss": 0.04821276814070834,
+          "test_ensemble_f1_score": 0.9764176126609346,
+          "test_ensemble_accuracy": 0.9822485207100592,
+          "test_cross_entropy": 0.0696995064404883,
+          "test_roc_auc": 0.9990603675985033,
+          "test_precision": 0.9822485207100592,
+          "test_recall": 0.9822485207100592,
+          "test_hinge_loss": 0.08216137762850832,
+          "test_f1_score": 0.9847006441834029,
+          "test_accuracy": 0.9822485207100592,
+          "best_hyperopt_params": {
+            "iterations": 100,
+            "depth": 3,
+            "learning_rate": 0.21030013684580467,
+            "l2_leaf_reg": 1.1158132462808803,
+            "random_strength": 8.456050055645859,
+            "bagging_temperature": 0.5071737962315873
+          },
+          "best_hyperopt_score": 0.9998691320499831,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.03464783399249427,
+          "fold": 3,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.030788229720661666,
+          "test_ensemble_roc_auc": 0.9989177489177489,
+          "test_ensemble_precision": 0.9881656804733728,
+          "test_ensemble_recall": 0.9881656804733728,
+          "test_ensemble_hinge_loss": 0.03458204484316483,
+          "test_ensemble_f1_score": 0.9680555555555556,
+          "test_ensemble_accuracy": 0.9881656804733728,
+          "test_cross_entropy": 0.030194228150793968,
+          "test_roc_auc": 0.9989177489177489,
+          "test_precision": 0.9822485207100592,
+          "test_recall": 0.9822485207100592,
+          "test_hinge_loss": 0.03541138452142673,
+          "test_f1_score": 0.9537223340040242,
+          "test_accuracy": 0.9822485207100592,
+          "best_hyperopt_params": {
+            "iterations": 500,
+            "depth": 4,
+            "learning_rate": 0.26547063984697133,
+            "l2_leaf_reg": 1.814953605240957,
+            "random_strength": 5.060497453317673,
+            "bagging_temperature": 0.13908816776105165
+          },
+          "best_hyperopt_score": 0.9999257504420209,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.19254129100590944,
+          "fold": 4,
           "imputed": false
         }
       ],
@@ -4174,7 +4703,14 @@
       "mean_cross_entropy": 0.060552971491056214,
       "mean_precision": 0.9822485207100592,
       "n_folds_imputed": 0,
-      "fully_imputed": false
+      "fully_imputed": false,
+      "mean_ensemble_accuracy": 0.9905325443786982,
+      "mean_ensemble_roc_auc": 0.9992089478318291,
+      "mean_ensemble_f1_score": 0.9841723551848368,
+      "mean_ensemble_precision": 0.9905325443786982,
+      "mean_ensemble_recall": 0.9905325443786982,
+      "mean_ensemble_cross_entropy": 0.08353015668986163,
+      "mean_ensemble_hinge_loss": 0.10841922708002723
     },
     "openml-469": {
       "folds": [
@@ -4189,6 +4725,133 @@
           "mem_usage": 8425.07421875,
           "time_usage": 0.4675540979988,
           "fold": 0,
+          "imputed": false,
+          "test_ensemble_cross_entropy": 1.7635684308103883,
+          "test_ensemble_roc_auc": 0.5856122353496468,
+          "test_ensemble_precision": 0.2125,
+          "test_ensemble_recall": 0.2125,
+          "test_ensemble_hinge_loss": 1.0296132876737425,
+          "test_ensemble_f1_score": 0.1821719116712698,
+          "test_ensemble_accuracy": 0.2125
+        },
+        {
+          "test_ensemble_cross_entropy": 1.7532687251994172,
+          "test_ensemble_roc_auc": 0.5882752829771228,
+          "test_ensemble_precision": 0.25,
+          "test_ensemble_recall": 0.25,
+          "test_ensemble_hinge_loss": 1.0372644946063463,
+          "test_ensemble_f1_score": 0.21180169037311894,
+          "test_ensemble_accuracy": 0.25,
+          "test_cross_entropy": 1.763308881168865,
+          "test_roc_auc": 0.5910541056388926,
+          "test_precision": 0.21875,
+          "test_recall": 0.21875,
+          "test_hinge_loss": 1.0233770104158606,
+          "test_f1_score": 0.15147484231991273,
+          "test_accuracy": 0.21875,
+          "best_hyperopt_params": {
+            "iterations": 500,
+            "depth": 7,
+            "learning_rate": 0.010573991487299012,
+            "l2_leaf_reg": 5.060136259064562,
+            "random_strength": 6.263471417802578,
+            "bagging_temperature": 0.17457459038297785
+          },
+          "best_hyperopt_score": 0.5848007215007216,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.4878037080052309,
+          "fold": 1,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 1.7707954136904511,
+          "test_ensemble_roc_auc": 0.6125298417334101,
+          "test_ensemble_precision": 0.225,
+          "test_ensemble_recall": 0.225,
+          "test_ensemble_hinge_loss": 1.0177491474476223,
+          "test_ensemble_f1_score": 0.1182784100945583,
+          "test_ensemble_accuracy": 0.225,
+          "test_cross_entropy": 1.7607287175139619,
+          "test_roc_auc": 0.5894617449149668,
+          "test_precision": 0.2125,
+          "test_recall": 0.2125,
+          "test_hinge_loss": 1.0249934713388185,
+          "test_f1_score": 0.1252777777777778,
+          "test_accuracy": 0.2125,
+          "best_hyperopt_params": {
+            "iterations": 500,
+            "depth": 7,
+            "learning_rate": 0.010863322344621458,
+            "l2_leaf_reg": 3.3074753553751766,
+            "random_strength": 6.708692242093393,
+            "bagging_temperature": 0.8877173639091734
+          },
+          "best_hyperopt_score": 0.5844582972582972,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.4756762499746401,
+          "fold": 2,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 1.7626361298292885,
+          "test_ensemble_roc_auc": 0.6279661200348099,
+          "test_ensemble_precision": 0.24375,
+          "test_ensemble_recall": 0.24375,
+          "test_ensemble_hinge_loss": 1.0178106283475308,
+          "test_ensemble_f1_score": 0.164496949727596,
+          "test_ensemble_accuracy": 0.24375,
+          "test_cross_entropy": 1.749451708359085,
+          "test_roc_auc": 0.6024701435166747,
+          "test_precision": 0.2,
+          "test_recall": 0.2,
+          "test_hinge_loss": 1.0417032552490473,
+          "test_f1_score": 0.18501684506429217,
+          "test_accuracy": 0.2,
+          "best_hyperopt_params": {
+            "iterations": 1000,
+            "depth": 7,
+            "learning_rate": 0.01092034624144032,
+            "l2_leaf_reg": 6.79537774444733,
+            "random_strength": 5.898534488498754,
+            "bagging_temperature": 0.8454827120592181
+          },
+          "best_hyperopt_score": 0.5818629437229437,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 1.6064482090005185,
+          "fold": 3,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 1.7739793228371634,
+          "test_ensemble_roc_auc": 0.5983419230939423,
+          "test_ensemble_precision": 0.20625,
+          "test_ensemble_recall": 0.20625,
+          "test_ensemble_hinge_loss": 1.01832792073398,
+          "test_ensemble_f1_score": 0.11544638935943284,
+          "test_ensemble_accuracy": 0.20625,
+          "test_cross_entropy": 1.7615293144058684,
+          "test_roc_auc": 0.5846705311444459,
+          "test_precision": 0.20625,
+          "test_recall": 0.20625,
+          "test_hinge_loss": 1.0279244916057528,
+          "test_f1_score": 0.17740620879779456,
+          "test_accuracy": 0.20625,
+          "best_hyperopt_params": {
+            "iterations": 100,
+            "depth": 5,
+            "learning_rate": 0.08190931585295598,
+            "l2_leaf_reg": 1.7670119622769058,
+            "random_strength": 9.168603290377627,
+            "bagging_temperature": 0.3637831881834725
+          },
+          "best_hyperopt_score": 0.5751801154401155,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.13220583298243582,
+          "fold": 4,
           "imputed": false
         }
       ],
@@ -4202,7 +4865,14 @@
       "mean_cross_entropy": 1.7547129065690263,
       "mean_precision": 0.20625,
       "n_folds_imputed": 0,
-      "fully_imputed": false
+      "fully_imputed": false,
+      "mean_ensemble_accuracy": 0.22749999999999998,
+      "mean_ensemble_roc_auc": 0.6025450806377863,
+      "mean_ensemble_f1_score": 0.15843907024519516,
+      "mean_ensemble_precision": 0.22749999999999998,
+      "mean_ensemble_recall": 0.22749999999999998,
+      "mean_ensemble_cross_entropy": 1.7648496044733417,
+      "mean_ensemble_hinge_loss": 1.0241530957618443
     },
     "openml-488": {
       "folds": [
@@ -16205,7 +16875,14 @@
           "mem_usage": 12763.0703125,
           "time_usage": 0.8168425289986772,
           "fold": 0,
-          "imputed": false
+          "imputed": false,
+          "test_ensemble_cross_entropy": 0.6057126139515093,
+          "test_ensemble_roc_auc": 0.6729164701319664,
+          "test_ensemble_precision": 0.822429906542056,
+          "test_ensemble_recall": 0.822429906542056,
+          "test_ensemble_hinge_loss": 0.4861196751015857,
+          "test_ensemble_f1_score": 0.11282051282051282,
+          "test_ensemble_accuracy": 0.822429906542056
         },
         {
           "test_ensemble_cross_entropy": 0.6155805632316902,
@@ -16339,13 +17016,13 @@
       "mean_precision": 0.8261682242990653,
       "n_folds_imputed": 0,
       "fully_imputed": false,
-      "mean_ensemble_accuracy": 0.8271028037383178,
-      "mean_ensemble_roc_auc": 0.5902883688143978,
-      "mean_ensemble_f1_score": 0.11580157782866275,
-      "mean_ensemble_precision": 0.8271028037383178,
-      "mean_ensemble_recall": 0.8271028037383178,
-      "mean_ensemble_cross_entropy": 0.6460947042057876,
-      "mean_ensemble_hinge_loss": 0.49209672015751504
+      "mean_ensemble_accuracy": 0.8261682242990653,
+      "mean_ensemble_roc_auc": 0.6068139890779116,
+      "mean_ensemble_f1_score": 0.11520536482703277,
+      "mean_ensemble_precision": 0.8261682242990653,
+      "mean_ensemble_recall": 0.8261682242990653,
+      "mean_ensemble_cross_entropy": 0.638018286154932,
+      "mean_ensemble_hinge_loss": 0.4909013111463292
     },
     "openml-44967": {
       "folds": [
@@ -16360,7 +17037,14 @@
           "mem_usage": 6393.3828125,
           "time_usage": 0.5065870639991772,
           "fold": 0,
-          "imputed": false
+          "imputed": false,
+          "test_ensemble_cross_entropy": 2.3438571874492036,
+          "test_ensemble_roc_auc": 0.6966569639972078,
+          "test_ensemble_precision": 0.20610687022900764,
+          "test_ensemble_recall": 0.20610687022900764,
+          "test_ensemble_hinge_loss": 1.0583257731227447,
+          "test_ensemble_f1_score": 0.054066500875011504,
+          "test_ensemble_accuracy": 0.20610687022900764
         },
         {
           "test_ensemble_cross_entropy": 2.277747910291964,
@@ -16494,13 +17178,13 @@
       "mean_precision": 0.1984732824427481,
       "n_folds_imputed": 0,
       "fully_imputed": false,
-      "mean_ensemble_accuracy": 0.1946564885496183,
-      "mean_ensemble_roc_auc": 0.6873957303406006,
-      "mean_ensemble_f1_score": 0.0590355713105203,
-      "mean_ensemble_precision": 0.1946564885496183,
-      "mean_ensemble_recall": 0.1946564885496183,
-      "mean_ensemble_cross_entropy": 2.285377343518036,
-      "mean_ensemble_hinge_loss": 1.0738039255511802
+      "mean_ensemble_accuracy": 0.19694656488549617,
+      "mean_ensemble_roc_auc": 0.6892479770719221,
+      "mean_ensemble_f1_score": 0.05804175722341853,
+      "mean_ensemble_precision": 0.19694656488549617,
+      "mean_ensemble_recall": 0.19694656488549617,
+      "mean_ensemble_cross_entropy": 2.2970733123042693,
+      "mean_ensemble_hinge_loss": 1.0707082950654931
     },
     "openml-44971": {
       "folds": [

results_academic_new/tabm.json

@@ -16189,17 +16189,278 @@
           "time_usage": 20.877522417002183,
           "fold": 3,
           "imputed": false
+        },
+        {
+          "test_cross_entropy": 0.5356540863525916,
+          "test_roc_auc": 0.9981818181818182,
+          "test_precision": 0.871875,
+          "test_recall": 0.871875,
+          "test_hinge_loss": 0.47088423371315,
+          "test_f1_score": 0.8632301587301587,
+          "test_accuracy": 0.871875,
+          "best_hyperopt_params": {
+            "lr": 0.0012193127638507766,
+            "weight_decay": 0.020619180487281848,
+            "dropout": 0.24873513547568987,
+            "d_embedding": 32
+          },
+          "best_hyperopt_score": 0.9978327721661054,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 18.420472290999896,
+          "fold": 4,
+          "imputed": false
+        }
+      ],
+      "mean_accuracy": 0.8737499999999999,
+      "mean_auc": 0.9987044051627384,
+      "mean_f1": 0.8667666666666667,
+      "mean_precision": 0.8737499999999999,
+      "mean_recall": 0.8737499999999999,
+      "mean_cross_entropy": 0.48715800647205515,
+      "mean_hinge_loss": 0.4478200674057007,
+      "mean_mem_usage": 0.0,
+      "mean_time_usage": 20.18519848319993,
+      "n_folds_imputed": 0,
+      "fully_imputed": false
+    },
+    "openml-1492": {
+      "folds": [
+        {
+          "test_cross_entropy": 1.156699615164213,
+          "test_roc_auc": 0.9910388608305275,
+          "test_precision": 0.6625,
+          "test_recall": 0.6625,
+          "test_hinge_loss": 0.7051838636398315,
+          "test_f1_score": 0.6486839826839828,
+          "test_accuracy": 0.6625,
+          "best_hyperopt_params": {
+            "lr": 0.002423325889994638,
+            "weight_decay": 0.022257666530381765,
+            "dropout": 0.38560332293301713,
+            "d_embedding": 12
+          },
+          "best_hyperopt_score": 0.9860106621773289,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 32.01086158299586,
+          "fold": 0,
+          "imputed": false
+        },
+        {
+          "test_cross_entropy": 1.2286058135066544,
+          "test_roc_auc": 0.9840144500561167,
+          "test_precision": 0.7,
+          "test_recall": 0.7,
+          "test_hinge_loss": 0.6779734492301941,
+          "test_f1_score": 0.6845331890331892,
+          "test_accuracy": 0.7,
+          "best_hyperopt_params": {
+            "lr": 0.00027373020261429836,
+            "weight_decay": 0.0001468268811765953,
+            "dropout": 0.4891702942631274,
+            "d_embedding": 8
+          },
+          "best_hyperopt_score": 0.9908512906846241,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 31.820702333003283,
+          "fold": 1,
+          "imputed": false
+        },
+        {
+          "test_cross_entropy": 1.156478408731966,
+          "test_roc_auc": 0.9899375701459036,
+          "test_precision": 0.6625,
+          "test_recall": 0.6625,
+          "test_hinge_loss": 0.6928892731666565,
+          "test_f1_score": 0.6502857142857142,
+          "test_accuracy": 0.6625,
+          "best_hyperopt_params": {
+            "lr": 0.0006248496919525067,
+            "weight_decay": 0.004066682696647132,
+            "dropout": 0.374137963929581,
+            "d_embedding": 12
+          },
+          "best_hyperopt_score": 0.982774410774411,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 29.472800083996844,
+          "fold": 2,
+          "imputed": false
+        },
+        {
+          "test_cross_entropy": 1.1330874912183488,
+          "test_roc_auc": 0.9901508136924803,
+          "test_precision": 0.7,
+          "test_recall": 0.7,
+          "test_hinge_loss": 0.6694453358650208,
+          "test_f1_score": 0.683531746031746,
+          "test_accuracy": 0.7,
+          "best_hyperopt_params": {
+            "lr": 0.0005569790535089744,
+            "weight_decay": 0.05060985454718844,
+            "dropout": 0.37690214509603104,
+            "d_embedding": 32
+          },
+          "best_hyperopt_score": 0.987324354657688,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 31.130173666009796,
+          "fold": 3,
+          "imputed": false
+        },
+        {
+          "test_cross_entropy": 1.0903145512050518,
+          "test_roc_auc": 0.9906530583613916,
+          "test_precision": 0.71875,
+          "test_recall": 0.71875,
+          "test_hinge_loss": 0.66400545835495,
+          "test_f1_score": 0.7036587301587303,
+          "test_accuracy": 0.71875,
+          "best_hyperopt_params": {
+            "lr": 0.00014892759435729974,
+            "weight_decay": 0.00010802075331235696,
+            "dropout": 0.45180888156939064,
+            "d_embedding": 32
+          },
+          "best_hyperopt_score": 0.9870695847362514,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 34.97199058299884,
+          "fold": 4,
+          "imputed": false
+        }
+      ],
+      "mean_accuracy": 0.68875,
+      "mean_auc": 0.9891589506172839,
+      "mean_f1": 0.6741386724386725,
+      "mean_precision": 0.68875,
+      "mean_recall": 0.68875,
+      "mean_cross_entropy": 1.1530371759652467,
+      "mean_hinge_loss": 0.6818994760513306,
+      "mean_mem_usage": 0.0,
+      "mean_time_usage": 31.881305649800925,
+      "n_folds_imputed": 0,
+      "fully_imputed": false
+    },
+    "openml-1493": {
+      "folds": [
+        {
+          "test_cross_entropy": 0.5716528967280577,
+          "test_roc_auc": 0.998027497194164,
+          "test_precision": 0.86875,
+          "test_recall": 0.86875,
+          "test_hinge_loss": 0.4614432454109192,
+          "test_f1_score": 0.8687698412698412,
+          "test_accuracy": 0.86875,
+          "best_hyperopt_params": {
+            "lr": 0.0018711960020417159,
+            "weight_decay": 0.0008478790047710117,
+            "dropout": 0.49444692904452914,
+            "d_embedding": 20
+          },
+          "best_hyperopt_score": 0.9972951739618405,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 19.16475325000647,
+          "fold": 0,
+          "imputed": false
+        },
+        {
+          "test_cross_entropy": 0.5297266848926852,
+          "test_roc_auc": 0.9980345117845119,
+          "test_precision": 0.84375,
+          "test_recall": 0.84375,
+          "test_hinge_loss": 0.43509262800216675,
+          "test_f1_score": 0.8399682539682541,
+          "test_accuracy": 0.84375,
+          "best_hyperopt_params": {
+            "lr": 0.0013613185775267942,
+            "weight_decay": 0.08148161871205739,
+            "dropout": 0.29946164105959094,
+            "d_embedding": 24
+          },
+          "best_hyperopt_score": 0.9970600448933782,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 19.984148166986415,
+          "fold": 1,
+          "imputed": false
+        },
+        {
+          "test_cross_entropy": 0.6539341132468874,
+          "test_roc_auc": 0.996210016835017,
+          "test_precision": 0.840625,
+          "test_recall": 0.840625,
+          "test_hinge_loss": 0.46700721979141235,
+          "test_f1_score": 0.8366587301587302,
+          "test_accuracy": 0.840625,
+          "best_hyperopt_params": {
+            "lr": 0.0002442290951725859,
+            "weight_decay": 0.0026316301528458335,
+            "dropout": 0.07481600311567849,
+            "d_embedding": 32
+          },
+          "best_hyperopt_score": 0.9978406285072952,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 20.21947816599277,
+          "fold": 2,
+          "imputed": false
+        },
+        {
+          "test_cross_entropy": 0.5036965041996426,
+          "test_roc_auc": 0.99765221661055,
+          "test_precision": 0.884375,
+          "test_recall": 0.884375,
+          "test_hinge_loss": 0.3905985951423645,
+          "test_f1_score": 0.8788174603174604,
+          "test_accuracy": 0.884375,
+          "best_hyperopt_params": {
+            "lr": 0.00011159732192444997,
+            "weight_decay": 0.00018611961079721345,
+            "dropout": 0.23031154209034704,
+            "d_embedding": 28
+          },
+          "best_hyperopt_score": 0.9963625140291807,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 20.636676833004458,
+          "fold": 3,
+          "imputed": false
+        },
+        {
+          "test_cross_entropy": 0.5794093607971741,
+          "test_roc_auc": 0.9979910213243545,
+          "test_precision": 0.85625,
+          "test_recall": 0.85625,
+          "test_hinge_loss": 0.46945613622665405,
+          "test_f1_score": 0.8486883116883119,
+          "test_accuracy": 0.85625,
+          "best_hyperopt_params": {
+            "lr": 0.0006746662057152432,
+            "weight_decay": 0.009468583756756063,
+            "dropout": 0.29612952877539017,
+            "d_embedding": 28
+          },
+          "best_hyperopt_score": 0.997199214365881,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 17.29036924999673,
+          "fold": 4,
+          "imputed": false
         }
       ],
-      "mean_accuracy": 0.8742187499999999,
-      "mean_auc": 0.9988350519079685,
-      "mean_f1": 0.8676507936507937,
-      "mean_precision": 0.8742187499999999,
-      "mean_recall": 0.8742187499999999,
-      "mean_cross_entropy": 0.47503398650192097,
-      "mean_hinge_loss": 0.44205402582883835,
+      "mean_accuracy": 0.85875,
+      "mean_auc": 0.9975830527497195,
+      "mean_f1": 0.8545805194805196,
+      "mean_precision": 0.85875,
+      "mean_recall": 0.85875,
+      "mean_cross_entropy": 0.5676839119728895,
+      "mean_hinge_loss": 0.4447195649147034,
       "mean_mem_usage": 0.0,
-      "mean_time_usage": 20.626380031249937,
+      "mean_time_usage": 19.45908513319737,
       "n_folds_imputed": 0,
       "fully_imputed": false
     }

results_academic_new/xgboost.json

@@ -3634,6 +3634,145 @@
           "mem_usage": 1224.171875,
           "time_usage": 0.2711171440023463,
           "fold": 0,
+          "imputed": false,
+          "test_ensemble_cross_entropy": 0.10579679117212816,
+          "test_ensemble_roc_auc": 0.8795131845841785,
+          "test_ensemble_precision": 0.0,
+          "test_ensemble_recall": 0.0,
+          "test_ensemble_hinge_loss": 1.0073315230537074,
+          "test_ensemble_f1_score": 0.0,
+          "test_ensemble_accuracy": 0.9704724409448819
+        },
+        {
+          "test_ensemble_cross_entropy": 0.08899522068489103,
+          "test_ensemble_roc_auc": 0.9235970250169034,
+          "test_ensemble_precision": 0.5,
+          "test_ensemble_recall": 0.06666666666666667,
+          "test_ensemble_hinge_loss": 1.0168359036845864,
+          "test_ensemble_f1_score": 0.11764705882352941,
+          "test_ensemble_accuracy": 0.9704724409448819,
+          "test_cross_entropy": 0.09432247165483593,
+          "test_roc_auc": 0.919134550371873,
+          "test_precision": 0.5,
+          "test_recall": 0.06666666666666667,
+          "test_hinge_loss": 1.0182012057263683,
+          "test_f1_score": 0.11764705882352941,
+          "test_accuracy": 0.9704724409448819,
+          "best_hyperopt_params": {
+            "n_estimators": 100,
+            "max_depth": 5,
+            "learning_rate": 0.023687007130470804,
+            "subsample": 0.8,
+            "colsample_bytree": 0.9925303442705216,
+            "gamma": 8.31521180051336,
+            "min_child_weight": 1,
+            "reg_alpha": 2.0271038233583633e-05,
+            "reg_lambda": 1.0282846803526135e-05
+          },
+          "best_hyperopt_score": 0.9169839255499154,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.08522262502810918,
+          "fold": 1,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.11408220546338878,
+          "test_ensemble_roc_auc": 0.9237322515212982,
+          "test_ensemble_precision": 0.0,
+          "test_ensemble_recall": 0.0,
+          "test_ensemble_hinge_loss": 1.021497095694171,
+          "test_ensemble_f1_score": 0.0,
+          "test_ensemble_accuracy": 0.9704724409448819,
+          "test_cross_entropy": 0.10086129943300651,
+          "test_roc_auc": 0.9112914131169709,
+          "test_precision": 0.0,
+          "test_recall": 0.0,
+          "test_hinge_loss": 1.0148767765920703,
+          "test_f1_score": 0.0,
+          "test_accuracy": 0.968503937007874,
+          "best_hyperopt_params": {
+            "n_estimators": 1000,
+            "max_depth": 8,
+            "learning_rate": 0.01905771627912354,
+            "subsample": 0.5,
+            "colsample_bytree": 0.9038950448163376,
+            "gamma": 2.3591617724605127,
+            "min_child_weight": 7,
+            "reg_alpha": 0.1242089740939823,
+            "reg_lambda": 0.10949804671547889
+          },
+          "best_hyperopt_score": 0.9131133671742809,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.5349197910109069,
+          "fold": 2,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.11126067959532937,
+          "test_ensemble_roc_auc": 0.8766734279918864,
+          "test_ensemble_precision": 0.3333333333333333,
+          "test_ensemble_recall": 0.13333333333333333,
+          "test_ensemble_hinge_loss": 1.013375514946482,
+          "test_ensemble_f1_score": 0.19047619047619047,
+          "test_ensemble_accuracy": 0.9665354330708661,
+          "test_cross_entropy": 0.10574140566116859,
+          "test_roc_auc": 0.8831643002028398,
+          "test_precision": 0.5,
+          "test_recall": 0.13333333333333333,
+          "test_hinge_loss": 1.016087757214064,
+          "test_f1_score": 0.21052631578947367,
+          "test_accuracy": 0.9704724409448819,
+          "best_hyperopt_params": {
+            "n_estimators": 100,
+            "max_depth": 3,
+            "learning_rate": 0.0828502753989465,
+            "subsample": 0.8,
+            "colsample_bytree": 0.7870945974375367,
+            "gamma": 5.10519210587262,
+            "min_child_weight": 2,
+            "reg_alpha": 0.005190784667518379,
+            "reg_lambda": 0.001044960081786356
+          },
+          "best_hyperopt_score": 0.8978003384094754,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.05686574999708682,
+          "fold": 3,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 0.10956683553416094,
+          "test_ensemble_roc_auc": 0.903448275862069,
+          "test_ensemble_precision": 0.16666666666666666,
+          "test_ensemble_recall": 0.06666666666666667,
+          "test_ensemble_hinge_loss": 1.015653877033367,
+          "test_ensemble_f1_score": 0.09523809523809523,
+          "test_ensemble_accuracy": 0.9625984251968503,
+          "test_cross_entropy": 0.11142064103339722,
+          "test_roc_auc": 0.8745098039215687,
+          "test_precision": 0.3333333333333333,
+          "test_recall": 0.13333333333333333,
+          "test_hinge_loss": 1.0149797739822686,
+          "test_f1_score": 0.19047619047619047,
+          "test_accuracy": 0.9665354330708661,
+          "best_hyperopt_params": {
+            "n_estimators": 100,
+            "max_depth": 5,
+            "learning_rate": 0.290892470443911,
+            "subsample": 1.0,
+            "colsample_bytree": 0.5065083677160951,
+            "gamma": 2.3290458323140326,
+            "min_child_weight": 3,
+            "reg_alpha": 0.00056919703494463,
+            "reg_lambda": 0.0005267841211909059
+          },
+          "best_hyperopt_score": 0.9125423011844334,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 0.04549491699435748,
+          "fold": 4,
           "imputed": false
         }
       ],
@@ -3647,7 +3786,14 @@
       "mean_hinge_loss": 1.017229706235734,
       "mean_precision": 0.0,
       "n_folds_imputed": 0,
-      "fully_imputed": false
+      "fully_imputed": false,
+      "mean_ensemble_accuracy": 0.9681102362204724,
+      "mean_ensemble_roc_auc": 0.9013928329952672,
+      "mean_ensemble_f1_score": 0.08067226890756303,
+      "mean_ensemble_precision": 0.2,
+      "mean_ensemble_recall": 0.05333333333333333,
+      "mean_ensemble_cross_entropy": 0.10594034648997966,
+      "mean_ensemble_hinge_loss": 1.0149387828824628
     },
     "openml-310": {
       "folds": [
@@ -7301,6 +7447,145 @@
           "mem_usage": 1680.4453125,
           "time_usage": 2.2674681289936416,
           "fold": 0,
+          "imputed": false,
+          "test_ensemble_cross_entropy": 1.2743783900908543,
+          "test_ensemble_roc_auc": 0.9958894500561166,
+          "test_ensemble_precision": 0.76875,
+          "test_ensemble_recall": 0.76875,
+          "test_ensemble_hinge_loss": 0.7816725173182006,
+          "test_ensemble_f1_score": 0.7457893217893218,
+          "test_ensemble_accuracy": 0.76875
+        },
+        {
+          "test_ensemble_cross_entropy": 2.5226355196814962,
+          "test_ensemble_roc_auc": 0.9952819865319865,
+          "test_ensemble_precision": 0.775,
+          "test_ensemble_recall": 0.775,
+          "test_ensemble_hinge_loss": 0.9523975096093942,
+          "test_ensemble_f1_score": 0.7621767676767677,
+          "test_ensemble_accuracy": 0.775,
+          "test_cross_entropy": 1.1078075419887774,
+          "test_roc_auc": 0.9956537598204265,
+          "test_precision": 0.76875,
+          "test_recall": 0.76875,
+          "test_hinge_loss": 0.7077239155769348,
+          "test_f1_score": 0.7567164502164503,
+          "test_accuracy": 0.76875,
+          "best_hyperopt_params": {
+            "n_estimators": 500,
+            "max_depth": 8,
+            "learning_rate": 0.027634675176833393,
+            "subsample": 0.8,
+            "colsample_bytree": 0.5822078211826398,
+            "gamma": 0.48373054658309955,
+            "min_child_weight": 5,
+            "reg_alpha": 0.005293087902580126,
+            "reg_lambda": 0.014100893399861808
+          },
+          "best_hyperopt_score": 0.9942485970819306,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 17.972933042008663,
+          "fold": 1,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 1.3729624794408313,
+          "test_ensemble_roc_auc": 0.9939751683501683,
+          "test_ensemble_precision": 0.7375,
+          "test_ensemble_recall": 0.7375,
+          "test_ensemble_hinge_loss": 0.8235238167134067,
+          "test_ensemble_f1_score": 0.7338968253968254,
+          "test_ensemble_accuracy": 0.7375,
+          "test_cross_entropy": 1.306754754622151,
+          "test_roc_auc": 0.9935851571268237,
+          "test_precision": 0.74375,
+          "test_recall": 0.74375,
+          "test_hinge_loss": 0.7987773418426514,
+          "test_f1_score": 0.7394624819624819,
+          "test_accuracy": 0.74375,
+          "best_hyperopt_params": {
+            "n_estimators": 1000,
+            "max_depth": 7,
+            "learning_rate": 0.03217510655252756,
+            "subsample": 0.8,
+            "colsample_bytree": 0.5039972131947854,
+            "gamma": 1.1923966662128533,
+            "min_child_weight": 7,
+            "reg_alpha": 0.03087459592333391,
+            "reg_lambda": 0.0021245647875177422
+          },
+          "best_hyperopt_score": 0.9952362514029179,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 27.33144712500507,
+          "fold": 2,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 2.9171989049592604,
+          "test_ensemble_roc_auc": 0.9948646184062852,
+          "test_ensemble_precision": 0.728125,
+          "test_ensemble_recall": 0.728125,
+          "test_ensemble_hinge_loss": 0.9694977487531868,
+          "test_ensemble_f1_score": 0.7185887445887447,
+          "test_ensemble_accuracy": 0.728125,
+          "test_cross_entropy": 1.2382730245151614,
+          "test_roc_auc": 0.9943946408529742,
+          "test_precision": 0.71875,
+          "test_recall": 0.71875,
+          "test_hinge_loss": 0.7639448046684265,
+          "test_f1_score": 0.7042411477411479,
+          "test_accuracy": 0.71875,
+          "best_hyperopt_params": {
+            "n_estimators": 1000,
+            "max_depth": 6,
+            "learning_rate": 0.18545650907804298,
+            "subsample": 1.0,
+            "colsample_bytree": 0.6902103832198208,
+            "gamma": 0.6626861547098757,
+            "min_child_weight": 7,
+            "reg_alpha": 5.04282745940483e-05,
+            "reg_lambda": 0.011530160289482628
+          },
+          "best_hyperopt_score": 0.9927890011223346,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 21.057493916974636,
+          "fold": 3,
+          "imputed": false
+        },
+        {
+          "test_ensemble_cross_entropy": 1.150709045212646,
+          "test_ensemble_roc_auc": 0.9950561167227833,
+          "test_ensemble_precision": 0.81875,
+          "test_ensemble_recall": 0.81875,
+          "test_ensemble_hinge_loss": 0.7441082015108526,
+          "test_ensemble_f1_score": 0.8050873015873015,
+          "test_ensemble_accuracy": 0.81875,
+          "test_cross_entropy": 1.1585127605016645,
+          "test_roc_auc": 0.9951199494949495,
+          "test_precision": 0.790625,
+          "test_recall": 0.790625,
+          "test_hinge_loss": 0.7472937703132629,
+          "test_f1_score": 0.7814444444444445,
+          "test_accuracy": 0.790625,
+          "best_hyperopt_params": {
+            "n_estimators": 1000,
+            "max_depth": 3,
+            "learning_rate": 0.1017580581507687,
+            "subsample": 0.5,
+            "colsample_bytree": 0.7692605340058155,
+            "gamma": 1.0289259111244764,
+            "min_child_weight": 5,
+            "reg_alpha": 1.1843418275523479e-05,
+            "reg_lambda": 0.23233587157122373
+          },
+          "best_hyperopt_score": 0.9949248035914703,
+          "metric_to_optimize": "roc_auc",
+          "mem_usage": 0.0,
+          "time_usage": 24.60832991701318,
+          "fold": 4,
           "imputed": false
         }
       ],
@@ -7314,7 +7599,14 @@
       "mean_hinge_loss": 0.9319642186164856,
       "mean_precision": 0.603125,
       "n_folds_imputed": 0,
-      "fully_imputed": false
+      "fully_imputed": false,
+      "mean_ensemble_accuracy": 0.765625,
+      "mean_ensemble_roc_auc": 0.9950134680134679,
+      "mean_ensemble_f1_score": 0.7531077922077922,
+      "mean_ensemble_precision": 0.765625,
+      "mean_ensemble_recall": 0.765625,
+      "mean_ensemble_cross_entropy": 1.8475768678770177,
+      "mean_ensemble_hinge_loss": 0.8542399587810081
     },
     "openml-1492": {
       "folds": [