Table 3 Optimization parameters and input features for the model
Methods | Features | Parameters |
|---|---|---|
SVM [19] | Raw(1-D) | C = 100, Degree = 10, Kernel = ‘rbf’ |
MLP [27] | Raw(1-D) | Neurons = 500, Hidden Layer = 5, Epochs = 200, Batch size = 5 |
CNN [34] | Raw(2-D) | Filters = 64, Dropout = 0.2, Hidden Layer = 7, Epochs = 180, Batch size = 10 |
KNN [20] | Raw (1-D) | Leaf size = 30, neighbours = 3, weight = ‘uniform’ |
LDA [25] | Raw(1-D) | n_components = 3, solver = 'svd' |
Proposed method | Raw (1-D and 2-D) | Filters = 32 & 64, Hidden Layer = (1 for wide component and 6 for deep Component), Lag = 11 day, Epochs = 100, Batch size = 5 |