Define the hyperparameters of the convolutional neural network CNN — param_CNN • cnnlstm

Define the hyperparameters of the convolutional neural network CNN

param_CNN(n_steps, n_filter, size_filter, pool_size, n_neuron)

Arguments

n_steps: Number of steps to split into samples CNN
n_filter: Number of filters in convolutional layer
size_filter: Size of filter in convolutional layer
pool_size: Size of pool in max_pooling layer
n_neuron: Number of neurons in dense layer

Value

Return CNN model architecture and model summary

Author

Catherine Rincon, catherine.rincon@udea.edu.co

Examples

# Example 1
# Generating a new CNN arquitecture

if (FALSE) {
library(keras)

model_CNN <- param_CNN(n_steps=6, n_filter=4,size_filter=2, pool_size=2, n_neuron=5)

summary(model_CNN)

}