13 #ifndef MLPACK_METHODS_ANN_LAYER_REPARAMETRIZATION_HPP
14 #define MLPACK_METHODS_ANN_LAYER_REPARAMETRIZATION_HPP
19 #include "../activation_functions/softplus_function.hpp"
53 typename InputDataType = arma::mat,
54 typename OutputDataType = arma::mat
56 class Reparametrization
71 const bool stochastic =
true,
72 const bool includeKl =
true,
73 const double beta = 1);
83 void Forward(
const arma::Mat<eT>& input, arma::Mat<eT>& output);
95 void Backward(
const arma::Mat<eT>& input,
96 const arma::Mat<eT>& gy,
105 OutputDataType
const&
Delta()
const {
return delta; }
107 OutputDataType&
Delta() {
return delta; }
120 return -0.5 * beta * arma::accu(2 * arma::log(stdDev) - arma::pow(stdDev, 2)
121 - arma::pow(mean, 2) + 1) / mean.n_cols;
131 double Beta()
const {
return beta; }
136 template<
typename Archive>
137 void serialize(Archive& ar,
const unsigned int );
153 OutputDataType delta;
156 OutputDataType gaussianSample;
163 OutputDataType preStdDev;
166 OutputDataType stdDev;
169 OutputDataType outputParameter;
176 #include "reparametrization_impl.hpp"
OutputDataType & Delta()
Modify the delta.
OutputDataType const & Delta() const
Get the delta.
The core includes that mlpack expects; standard C++ includes and Armadillo.
void Forward(const arma::Mat< eT > &input, arma::Mat< eT > &output)
Ordinary feed forward pass of a neural network, evaluating the function f(x) by propagating the activ...
bool Stochastic() const
Get the value of the stochastic parameter.
double Loss()
Get the KL divergence with standard normal.
void serialize(Archive &ar, const unsigned int)
Serialize the layer.
bool IncludeKL() const
Get the value of the includeKl parameter.
OutputDataType & OutputParameter()
Modify the output parameter.
void Backward(const arma::Mat< eT > &input, const arma::Mat< eT > &gy, arma::Mat< eT > &g)
Ordinary feed backward pass of a neural network, calculating the function f(x) by propagating x backw...
OutputDataType const & OutputParameter() const
Get the output parameter.
size_t & OutputSize()
Modify the output size.
Reparametrization()
Create the Reparametrization object.
size_t const & OutputSize() const
Get the output size.
double Beta() const
Get the value of the beta hyperparameter.