List your company
ResearchFormerly WizWand

Sepsis Prediction on eICU

98.2AUROCCatB
81.66485.95790.2594.543Apr 2, 2026
Updated 7mo ago

Evaluation results

MethodLinks
CatBModel Category=Classical and Ensemble Models2026.0498.2—76
LightGModel Category=Classical and Ensemble Models2026.0498.2—75.7
XGBModel Category=Classical and Ensemble Models2026.0498.2—76.2
TabMModel Category=Deep Tabular Models2026.0498.2—75.5
TromptModel Category=Deep Tabular Models2026.0497.9—73.9
kNNPFNModel Category=Ours, AWARE=true2026.0497.9—73.9
ExcelFModel Category=Deep Tabular Models2026.0497.6—67.3
MNCAModel Category=Deep Tabular Models2026.0497.5—72.3
TabDPTModel Category=Ours, AWARE=true2026.0497.2—70
ResNetModel Category=Deep Tabular Models2026.0497.1—70.3
kNNPFNModel Category=Tabular In-Context Learning Models2026.0496.9—69.1
TabPFNModel Category=Tabular In-Context Learning Models2026.0496.8—69.5
MLPModel Category=Classical and Ensemble Models2026.0496.5—66.5
KNNModel Category=Ours, AWARE=true2026.0496.3—69.5
TabDPTModel Category=Tabular In-Context Learning Models2026.0496—65.7
TabNetModel Category=Deep Tabular Models2026.0495.9—63.1
TabTModel Category=Deep Tabular Models2026.0495.5—61.4
LRModel Category=Classical and Ensemble Models2026.0492.9—54
KNNModel Category=Classical and Ensemble Models2026.0482.3—30.3
cs-CoxEvaluation=5-fold cross-validation2025.12—0.065—
DeepHitEvaluation=5-fold cross-validation2025.12—0.058—
DeepSurvEvaluation=5-fold cross-validation2025.12—0.064—
DynamicDeepHitEvaluation=5-fold cross-validation2025.12—0.058—
SurvTRACEEvaluation=5-fold cross-validation2025.12—0.065—
TRACERCovariate Embedding with Time-Decay=false, Factorized Attention=true, Evaluation=5-fold cross-validation2025.12—0.056—
TRACERCovariate Embedding with Time-Decay=true, Factorized Attention=false, Evaluation=5-fold cross-validation2025.12—0.055—
TRACERCovariate Embedding with Time-Decay=true, Factorized Attention=true, Evaluation=5-fold cross-validation2025.12—0.055—