Calibration on Tabular datasets benchmark leaderboard · SOTA2 Research

Calibration on Tabular datasets

0.2983NLLPlatt

Evaluation Results

Method	Links
Platt2025.10		0.2983	82.44	0.0897	2.66	66.2
Platt2025.10		0.2983	82.32	0.0897	2.57	57.25
Beta2025.10		0.2986	82.31	0.0898	2.59	57.1
Platt2025.10		0.2988	82.15	0.0897	2.61	73.18
Beta2025.10		0.2989	82.17	0.0898	2.71	73.85
Beta2025.10		0.299	82.31	0.09	2.76	66.72
Beta2025.10		0.2997	82.08	0.0898	2.59	57.76
Platt2025.10		0.2998	82.07	0.0898	2.56	57.7
Histogram Binning2025.10		0.3044	79.01	0.0909	2.74	53.81
Binwise-Beta2025.10		0.3074	80.89	0.0919	2.94	56.92
Binwise-Platt2025.10		0.3076	80.81	0.092	2.95	56.8
Bayesian Binning2025.10		0.3109	78.56	0.0919	2.91	55.93
Binwise-TS2025.10		0.324	80.97	0.0995	8.28	59.07
LoRe2025.10		0.3256	80.99	0.0946	3.57	58.73
TS2025.10		0.341	82.21	0.1066	11.91	68.56
Isotonic2025.10		0.3422	81.75	0.0909	2.83	56.54
TS2025.10		0.3425	82.19	0.1071	12.08	65.53
TS2025.10		0.3441	82.11	0.1075	12.21	75.42
ensemble-TS2025.10		0.3447	82.08	0.1076	12.23	65.43
TS2025.10		0.3456	82.08	0.1077	12.21	65.43
base2025.10		0.3756	82.08	0.1104	13.03	67.36