MNIST::PositiveWeights Class Reference

Constraint for weights in neural network: Only weights >0. More...

#include <mnist_mlp.hpp>

Public Member Functions
void	setup (std::vector< cypress::Matrix< Real >> &)
void	constrain_weights (std::vector< cypress::Matrix< Real >> &layers)

Detailed Description

Constraint for weights in neural network: Only weights >0.

Definition at line 148 of file mnist_mlp.hpp.

Member Function Documentation

void MNIST::PositiveWeights::constrain_weights ( std::vector< cypress::Matrix< Real >> &  layers )

inline

Definition at line 151 of file mnist_mlp.hpp.

    {
        for (auto &i : layers) {
            for (auto &j : i) {
                if (j < 0.0) {
                    j = 0.0;
                }
            }
        }
    }

void MNIST::PositiveWeights::setup ( std::vector< cypress::Matrix< Real >> &  )

inline

Definition at line 150 of file mnist_mlp.hpp.

{}

---

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

• source/SNABs/mnist/mnist_mlp.hpp