nn::Linear Class Reference

Linear module. This implementation mostly follows PyTorch's implementation. More...

#include <nn.h>

Public Member Functions
	Linear (int, int, bool bias=false)
	Default constructor for nn::Linear class.
std::vector< double >	forward (std::vector< double > const &)
	forward function: Forwards an input 1D tensor to the network
void	update_weights (std::vector< double > const &, double, double)
	updates the weight vector by value

Detailed Description

Linear module. This implementation mostly follows PyTorch's implementation.

Constructor & Destructor Documentation

Linear()

nn::Linear::Linear ( int in_features, int out_features, bool bias = false )

inlineexplicit

Default constructor for nn::Linear class.

Parameters

in_features(int)	The input features
out_features(int)	The output features
bias(bool)	If set to true, then bias will be initialized with a uniform distribution on U(-1.0, 1.0)

Member Function Documentation

forward()

std::vector< double > nn::Linear::forward ( std::vector< double > const & input_tensor )

inline

forward function: Forwards an input 1D tensor to the network

Parameters

input_tensor

1D vector, the input tensor

Returns

1D vector(wT * x + bias)

update_weights()

void nn::Linear::update_weights ( std::vector< double > const & input, double error, double learning_rate )

inline

updates the weight vector by value

Parameters

value

double, the value that will be added to weight vector

---

The documentation for this class was generated from the following file:

• /Users/runner/work/AlgoPlus/AlgoPlus/src/machine_learning/nn/nn.h