Table 5.

Confusion matrix for different models in training and test sets

Prediction	Dataset	Model	Major effect	Minor effect	Moderate effect
Major effect	Training	LGBM	1530	74	2
		RF	587	87	2
		LR	969	150	5
	Test	LGBM	575	105	3
		RF	1342	265	6
		LR	97	186	8
Minor effect	Training	LGBM	123	1446	34
		RF	66	573	36
		LR	177	851	125
	Test	LGBM	108	456	120
		RF	240	1133	239
		LR	202	766	166
Moderate effect	Training	LGBM	29	102	1464
		RF	14	74	620
		LR	48	238	869
	Test	LGBM	28	156	509
		RF	72	376	1132
		LR	58	272	803

LGBM=Light gradient boosting machine, LR=Logistic regression, RF=Random forest