Table 2 Main parameters of the SVM, LSTM, ResNet18, VGG16 and AlexNet models
From: A deep learning model for rapid classification of tea coal disease
Model | Model parameters | Value | |
|---|---|---|---|
SVM | The Kernel Function | Polynomial Kernel | |
Cache_size | 200 | ||
Tol (Tolerance Used in the IterativeAlgorithm) | 10− 3 | ||
Max_iter | -1 | ||
C (Regularization Parameter) | 1 | ||
LSTM | Normalize | L2 | |
Optimizer | Adam (Adaptive moment estimation) | ||
Activation Function | Tanh (TanHyperbolic) | ||
NumHiddenUnits | 20 | ||
Learning Rate | 0.001 | ||
Epochs | 40 | ||
Batch Size | 64 | ||
Dropout | 0.5 | ||
Verbose | 1 | ||
ResNet18 | Learning Rate | 0.001 | |
Epochs | 90 | ||
Batch Size | 32 | ||
Activation Function | ReLU | ||
Normalize | L2 | ||
VGG16 | Learning Rate | 0.001 | |
Epochs | 90 | ||
Batch Size | 16 | ||
Activation Function | ReLU | ||
Normalize | L2 | ||
AlexNet | Learning Rate | 0.001 | |
Epochs | 90 | ||
Batch Size | 16 | ||
Activation Function | ReLU | ||
Normalize | L2 | ||