Skip to main content

View full-text article in PMC

. 2022 Oct 17;13:890221. doi: 10.3389/fphar.2022.890221

TABLE 2.

Prediction performance of different algorithms.

Metrics algorithms	Dose regimen ^a	Precision	Recall	f1_score	Accuracy	AUC	Sensitivity	Specificity
XGBoost	0	0.45	0.83	0.59	0.79	0.87	0.78	0.83
XGBoost	1	0.95	0.78	0.86	0.79	0.87	0.78	0.83
LightGBM	0	0.50	0.67	0.57	0.82	0.84	0.85	0.67
LightGBM	1	0.92	0.85	0.88	0.82	0.84	0.85	0.67
CatBoost	0	0.56	0.83	0.67	0.85	0.91	0.85	0.83
CatBoost	1	0.96	0.85	0.90	0.85	0.91	0.85	0.83
Random forest	0	0.40	0.67	0.50	0.76	0.88	0.78	0.67
Random forest	1	0.91	0.78	0.84	0.76	0.88	0.78	0.67
GBDT	0	0.29	0.33	0.31	0.73	0.77	0.81	0.33
GBDT	1	0.85	0.81	0.83	0.73	0.77	0.81	0.33
SVM	0	0.44	0.67	0.53	0.79	0.82	0.81	0.67
SVM	1	0.92	0.81	0.86	0.79	0.82	0.81	0.67
Logistic regression	0	0.44	0.67	0.53	0.79	0.81	0.81	0.67
Logistic regression	1	0.92	0.81	0.86	0.79	0.81	0.81	0.67
ANN	0	0.40	0.67	0.50	0.76	0.85	0.78	0.67
ANN	1	0.91	0.78	0.84	0.76	0.85	0.78	0.67
TabNet	0	0.44	0.67	0.53	0.79	0.8	0.81	0.67
TabNet	1	0.92	0.81	0.86	0.79	0.8	0.81	0.67

^{^a}

Regimen of the daily dose of 0.5 g VPA corresponds to “0,” and regimen of the daily dose of 1 g VPA corresponds to “1”.