Table 11.

The performance of ensemble model classification per feature for all classifiers.

Feature	All three models ensemble
	ACC	Prec	Recall	F1	Spec	NPV	kappa	AUC
SPEC	0.73	0.70	0.80	0.75	0.66	0.77	0.46	0.73
Chroma	0.63	0.67	0.50	0.57	0.75	0.60	0.25	0.63
MFCC	0.67	0.69	0.61	0.65	0.73	0.65	0.34	0.67
MelSpectrum	0.74	0.72	0.78	0.75	0.70	0.76	0.48	0.74
PowerSPEC	0.67	0.71	0.56	0.63	0.77	0.64	0.34	0.67
RAW	0.59	0.60	0.50	0.55	0.67	0.57	0.17	0.59
Tonal	0.54	0.54	0.53	0.53	0.54	0.53	0.07	0.54
All features	0.63	0.61	0.76	0.67	0.51	0.68	0.26	0.63

Bold value indicates best performance of the classifiers.