Table 2 The parameters of CNN model
From: Classification of lung cancer severity using gene expression data based on deep learning
Parameter | Value | Meaning |
|---|---|---|
Random Seeds | 7 | To fix the reproducibility of the results over multiple runs of the code. |
No. of epochs | 50 | The number of iterations of all the training dataset. |
Activation function of the hidden layer | RELU | The default activation function for the hidden layers in CNN. |
Activation function of the output layer | Sigmoid | It is a binary classification activation function of the output layer. |
Patience value | 5 | It refers to the number of epochs the model can stop the training process, if there are no improvements. |
Optimizer | Adam | It refers to Adaptive movement estimation algorithm that is utilized to update the network weight iterative based in training data. |
Loss function | Binary cross entropy | It is used to compare the actual label with the predicted output. |
Filter size | 16 | The number of channels in the output of the convolutional layer. |
Kernel size | 5 | The size of the convolutional filters. |