Utility of support vector machine and decision tree to identify the prognosis of metformin poisoning in the United States: analysis of National Poisoning Data System

Table 2 Characteristics for the training and test sets of SVM and DT

Labels	Dataset	Model	Major effect	Minor effect	Moderate effect	Average	Weighted average
Specificity	Training Set	DT	0.990123	0.798817	0.938451	0.909130	0.856677
	Training Set	SVM	0.999506	0.784615	0.961272	0.915131	0.856145
	Test Set	DT	0.983752	0.765734	0.941300	0.896929	0.838008
	Test Set	SVM	0.986706	0.786713	0.947589	0.907003	0.852953
Precision	Training Set	DT	0.807692	0.881616	0.855987	0.848432	0.868604
	Training Set	SVM	0.989247	0.876275	0.905724	0.923749	0.892954
	Test Set	DT	0.717949	0.862140	0.856410	0.812166	0.851595
	Test Set	SVM	0.750000	0.874227	0.874372	0.832866	0.866857
Recall	Training Set	DT	0.631579	0.964204	0.742978	0.779587	0.870714
	Training Set	SVM	0.691729	0.981721	0.755618	0.809690	0.889249
	Test Set	DT	0.651163	0.965438	0.687243	0.767948	0.852778
	Test Set	SVM	0.627907	0.976959	0.716049	0.773638	0.868056
F1_score	Training Set	DT	0.708861	0.921062	0.795489	0.808471	0.866553
	Training Set	SVM	0.814159	0.926006	0.823890	0.854685	0.885421
	Test Set	DT	0.682927	0.910870	0.762557	0.785451	0.847201
	Test Set	SVM	0.683544	0.922742	0.787330	0.797872	0.862755
Accuracy	Training Set	DT	NaN	NaN	NaN	0.870714	0.870714
	Training Set	SVM	NaN	NaN	NaN	0.889249	0.889249
	Test Set	DT	NaN	NaN	NaN	0.852778	0.852778
	Test Set	SVM	NaN	NaN	NaN	0.868056	0.868056

ISSN: 2050-6511