#include <fc_layer.h>
◆ Fc_Layer() [1/2]
Fc_Layer::Fc_Layer |
( |
int |
input_size, |
|
|
int |
output_size |
|
) |
| |
Constructor of the Full connected Layer, Generate Weight & Bias random Matrix
- Parameters
-
input_size | size of rows
|
output_size | size of cols
|
◆ Fc_Layer() [2/2]
Neural::Fc_Layer::Fc_Layer |
( |
Eigen::MatrixXd |
weights, |
|
|
Eigen::MatrixXd |
bias |
|
) |
| |
◆ Backward_propagation()
MatrixXd Fc_Layer::Backward_propagation |
( |
Eigen::MatrixXd |
output_error, |
|
|
float |
learning_rate |
|
) |
| |
|
virtual |
Performs retro propagation on current layer
- Parameters
-
output_error | The inputs of the Layer = The outputs of the previous Layer, or The data of the first Layer |
learning_rate | The step size at each iteration |
- Returns
- Matrix of input Layer error /!\ -> Matrix stored row or col
Implements Neural::Layer.
◆ Forward_propagation()
MatrixXd Fc_Layer::Forward_propagation |
( |
Eigen::MatrixXd |
input | ) |
|
|
virtual |
Performs forward propagation on the current layer
- Parameters
-
input_data | The inputs of the Layer = The outputs of the previous Layer, or The data of the first Layer |
- Returns
- Output Matrix of forward propagation results
Implements Neural::Layer.
◆ GetBias()
MatrixXd Fc_Layer::GetBias |
( |
| ) |
|
◆ GetWeights()
MatrixXd Fc_Layer::GetWeights |
( |
| ) |
|
◆ SetBias()
void Fc_Layer::SetBias |
( |
Eigen::MatrixXd |
weights | ) |
|
◆ SetWeights()
void Fc_Layer::SetWeights |
( |
Eigen::MatrixXd |
weights | ) |
|
◆ toJSON()
Json::Value Fc_Layer::toJSON |
( |
| ) |
|
|
virtual |
◆ m_bias
Eigen::MatrixXd Neural::Fc_Layer::m_bias |
|
protected |
◆ m_weights
Eigen::MatrixXd Neural::Fc_Layer::m_weights |
|
protected |
The documentation for this class was generated from the following files: