Table 5 Accuracy scores (in %) comparison of spectrograms and hybrid models.

ML classifiers	\(\:{S}_{M}\)			\(\:{S}_{C}\)			\(\:{S}_{G}\)
ML classifiers	MobileNetV2	Inception V3	ResNet50	MobileNetV2	Inception V3	ResNet50	MobileNetV2	Inception V3	ResNet50
Random forest	89.47	83.33	93.49	81.58	78.07	80.70	78.95	78.95	85.09
SVM	93.86	89.47	95.37	82.46	86.24	83.33	87.72	86.84	89.47
KNN	94.74	85.96	92.11	73.68	69.30	76.32	85.09	79.82	78.07
LMT	95.59	92.11	96.49	85.09	84.21	87.72	85.09	89.35	90.35
Naive Bayes	71.93	78.07	70.18	56.14	62.28	50.00	54.39	63.16	59.65
Linear SVM	95.61	92.98	94.37	81.58	86.47	85.96	90.35	88.60	87.72
MLP	92.11	88.60	96.25	83.33	85.09	82.46	85.09	87.72	89.47
Decision tree	60.53	63.16	84.21	48.25	57.02	58.77	55.26	52.63	58.77
Gradient booster	83.33	83.33	90.35	71.05	78.07	78.95	67.54	64.91	78.07
Adaboost	24.56	22.81	44.74	24.56	28.07	31.58	31.58	24.56	24.56

Quick links

Search