A robust Parkinson’s disease detection model based on time-varying synaptic efficacy function in spiking neural network

Table 6 Performance comparison for K-fold cross validation with different values of K for dataset #2

Technology	Accuracy (in %)	Sensitivity (in %)	Specificity (in %)	MCC	Precision	F1 Score	Gmean
K = 3
MLP-NN	78.45	81.23	74.65	0.5612	0.8023	0.8345	0.7864
RBF-NN	81.27	76.54	85.12	0.6138	0.8432	0.8117	0.798
RNN	80.02	80.55	79.42	0.5914	0.8056	0.8015	0.8002
LSTM	79.03	73.12	84.78	0.5784	0.7986	0.7623	0.7664
SEFRON [Dataset#2]	86.12	85.23	87.34	0.7325	0.8724	0.8506	0.8652
K = 5
MLP-NN	77.88	75.34	78.56	0.5431	0.791	0.7709	0.7736
RBF-NN	84.11	90.67	77.25	0.6659	0.8154	0.8523	0.8459
RNN	84.03	82.47	86.65	0.7199	0.8412	0.835	0.8325
LSTM	81.67	78.22	82.4	0.6197	0.8061	0.8305	0.8183
SEFRON [Dataset#2]	88.03	86.12	89.85	0.7551	0.9023	0.8825	0.8861
K = 8
MLP-NN	82.14	78.67	85.42	0.6115	0.8189	0.8027	0.8049
RBF-NN	86.32	81.25	92.1	0.7424	0.9057	0.8615	0.8722
RNN	84.03	82.47	86.65	0.7199	0.8412	0.835	0.8325
LSTM	83.09	80.21	85.18	0.6743	0.795	0.8202	0.8111
SEFRON [Dataset#2]	90.14	89.45	91.32	0.7921	0.915	0.8913	0.8957
K = 10
MLP-NN	83.33	79.32	88.67	0.6478	0.85	0.81	0.816
RBF-NN	85.22	84.45	90.13	0.7115	0.9	0.87	0.877
RNN	86	88.77	86.5	0.7521	0.873	0.8714	0.8750
LSTM	84.53	86.67	82.22	0.6814	0.7905	0.8327	0.8141
SEFRON [Dataset#2]	91.94	99.95	87.69	0.82	0.77	0.89	0.87

ISSN: 1471-2377