Table 3.

Performance comparison between pipelined classifier models.

Pipeline models	Features	Accuracy (%)	Accuracy deviation	F1-Score (%)	Precision (%)	Recall (%)	MCC (%)
PCA + SVC (linear)	31	96	0.01	96	95	96	91
PCA + SVC (RBF)	34	92	0.02	92	91	92	86
PCA + Decision Tree (Gini)	18	75	0.03	75	74	76	71
PCA + Decision Tree (Entropy)	27	76	0.02	76	75	76	72
PCA + Naive Bayes	36	91	0.02	91	90	91	85
PCA + KNN	21	82	0.02	82	80	83	76