Visit-level Disease Prediction (10 diseases) on MIMIC III (1000 index visits) benchmark leaderboard · SOTA2 Research

Visit-level Disease Prediction (10 diseases) on MIMIC III (1000 index visits)

0.854AccuracyLLaMA3-8B + KG-CoT

Evaluation Results

Method	Links
LLaMA3-8B + KG-CoT2025.12		0.854	0.6966	0.4662	0.4053
Gemma-7B + KG-CoT2025.12		0.8458	0.6609	0.4363	0.4263
XGBoost2025.12		0.8159	0.6799	0.4309	0.3065
RF2025.12		0.8107	0.6574	0.4024	0.2191
Gemma-7B2025.12		0.7475	0.5152	0.2528	0.0169
SVM2025.12		0.7391	0.6873	0.4256	0.1275
Logistic Regression2025.12		0.728	0.6874	0.4178	0.1238
MLP2025.12		0.7029	0.6642	0.3911	0.1917
SGD2025.12		0.6886	0.6563	0.384	0.1508
Naive Bayes2025.12		0.5394	0.6224	0.3581	0.3836
LLaMA3-8B2025.12		0.4709	0.5604	0.2773	0.3327