Model

Hyperparameters

Random Forest (RF)

Number of Estimators: 400

Minimum Samples Split: 5

Minimum Samples Leaf: 1

Maximum Features: “sqrt”

Maximum Depth: 30

Bootstrap: True

Convolutional Neural Network (CNN)

Optimizer: Adam

Learning Rate: 0.001

Epochs: 50

Dropout Rate: 0.5

Dense Units: 256

Dense Layers: 2

Convolutional Layers: 1

Convolutional Kernel Size: 5

Convolutional Filters: 128

Batch Size: 64

Activation Function: ReLU

Hybrid Model

Base Estimator Maximum Depth: 7

Learning Rate: 0.01

Number of Estimators: 200