Data Clustering Improves Siamese Neural Networks Classification of Parkinson’s Disease

Table 2

A comparison between applying the proposed model on Parkinson dataset without clustering and clustered PD dataset with the performance metrics accuracy, precision, recall, specificity, and f₁_score.

Class	Parkinson dataset without clustering					Clustered Parkinson dataset
	Accuracy (%)	Precision (%)	Recall (%)	Specificity (%)	F1_score (%)	Accuracy (%)	Precision (%)	Recall (%)	Specificity (%)	F1_score (%)
PD	71.88	56.76	65.63	75	60.87	77.05	50	42.86	87.23	46.15
RBD	72.92	57.89	68.75	75	62.86	75.36	60.87	63.64	80.85	62.22
HC	92.71	100	78.13	100	87.72	100	100	100	100	100
Weighted average	79.17	71.55	70.83	83.33	71.19	85.4	73.06	72	90.06	72.53