Table 7 Ablation study results.

From: Enhancing parkinson disease detection through feature based deep learning with autoencoders and neural networks

Configuration

Accuracy (%)

Precision (%)

Recall (%)

F1 Score (%)

Full Model (FB-DNN)

96.15

98

98

98

Without Stacked Autoencoder (SAE)

89.5

91

90.5

90.75

Without Modified Band Pass Filter

87.3

89.1

88.4

88.75

Without Regularization Techniques

85.7

86.9

86.1

86.5

Raw Audio Features + DNN Only

83.25

84.1

82.8

83.45

Back to article page