Logo ROOT   master
Reference Guide
Functions.h File Reference
This graph shows which files directly or indirectly include this file:

Namespaces

 TMVA
 create variable transformations
 
 TMVA::DNN
 

Enumerations

enum  TMVA::DNN::EActivationFunction {
  TMVA::DNN::EActivationFunction::kIdentity = 0, TMVA::DNN::EActivationFunction::kRelu = 1, TMVA::DNN::EActivationFunction::kSigmoid = 2, TMVA::DNN::EActivationFunction::kTanh = 3,
  TMVA::DNN::EActivationFunction::kSymmRelu = 4, TMVA::DNN::EActivationFunction::kSoftSign = 5, TMVA::DNN::EActivationFunction::kGauss = 6, TMVA::DNN::EActivationFunction::kFastTanh = 7
}
 Enum that represents layer activation functions. More...
 
enum  TMVA::DNN::EInitialization {
  TMVA::DNN::EInitialization::kGauss = 'G', TMVA::DNN::EInitialization::kUniform = 'U', TMVA::DNN::EInitialization::kIdentity = 'I', TMVA::DNN::EInitialization::kZero = 'Z',
  TMVA::DNN::EInitialization::kGlorotNormal = 'X', TMVA::DNN::EInitialization::kGlorotUniform = 'F'
}
 
enum  TMVA::DNN::ELossFunction { TMVA::DNN::ELossFunction::kCrossEntropy = 'C', TMVA::DNN::ELossFunction::kMeanSquaredError = 'R', TMVA::DNN::ELossFunction::kSoftmaxCrossEntropy = 'S' }
 Enum that represents objective functions for the net, i.e. More...
 
enum  TMVA::DNN::EOptimizer {
  TMVA::DNN::EOptimizer::kSGD = 0, TMVA::DNN::EOptimizer::kAdam = 1, TMVA::DNN::EOptimizer::kAdagrad = 2, TMVA::DNN::EOptimizer::kRMSProp = 3,
  TMVA::DNN::EOptimizer::kAdadelta = 4
}
 Enum representing the optimizer used for training. More...
 
enum  TMVA::DNN::EOutputFunction { TMVA::DNN::EOutputFunction::kIdentity = 'I', TMVA::DNN::EOutputFunction::kSigmoid = 'S', TMVA::DNN::EOutputFunction::kSoftmax = 'M' }
 Enum that represents output functions. More...
 
enum  TMVA::DNN::ERegularization { TMVA::DNN::ERegularization::kNone = '0', TMVA::DNN::ERegularization::kL1 = '1', TMVA::DNN::ERegularization::kL2 = '2' }
 Enum representing the regularization type applied for a given layer. More...
 

Functions

template<typename Architecture_t >
void TMVA::DNN::addRegularizationGradients (typename Architecture_t::Matrix_t &A, const typename Architecture_t::Matrix_t &W, typename Architecture_t::Scalar_t weightDecay, ERegularization R)
 Add the regularization gradient corresponding to weight matrix W, to the matrix A. More...
 
template<typename Architecture_t >
void TMVA::DNN::evaluate (typename Architecture_t::Tensor_t &A, EActivationFunction f)
 Apply the given activation function to each value in the given tensor A. More...
 
template<typename Architecture_t >
void TMVA::DNN::evaluate (typename Architecture_t::Matrix_t &A, EOutputFunction f, const typename Architecture_t::Matrix_t &X)
 Apply the given output function to each value in the given tensor A. More...
 
template<typename Architecture_t >
auto TMVA::DNN::evaluate (ELossFunction f, const typename Architecture_t::Matrix_t &Y, const typename Architecture_t::Matrix_t &output, const typename Architecture_t::Matrix_t &weights) -> decltype(Architecture_t::CrossEntropy(Y, output, weights))
 Compute the value of the objective function f for given activations of the ouput layer and the truth Y. More...
 
template<typename Architecture_t >
void TMVA::DNN::evaluateDerivative (typename Architecture_t::Tensor_t &B, EActivationFunction f, const typename Architecture_t::Tensor_t &A)
 Compute the first partial derivative of the activation function for the values given in tensor A and write the results into B. More...
 
template<typename Architecture_t >
void TMVA::DNN::evaluateDerivativeMatrix (typename Architecture_t::Matrix_t &B, EActivationFunction f, const typename Architecture_t::Matrix_t &A)
 
template<typename Architecture_t >
void TMVA::DNN::evaluateGradients (typename Architecture_t::Matrix_t &dY, ELossFunction f, const typename Architecture_t::Matrix_t &Y, const typename Architecture_t::Matrix_t &output, const typename Architecture_t::Matrix_t &weights)
 Compute the gradient of the given output function f for given activations output of the output layer and truth Y and write the results into dY. More...
 
template<typename Architecture_t >
void TMVA::DNN::evaluateMatrix (typename Architecture_t::Matrix_t &A, EActivationFunction f)
 
template<typename Architecture_t >
void TMVA::DNN::initialize (typename Architecture_t::Matrix_t &A, EInitialization m)
 
template<typename Architecture_t >
auto TMVA::DNN::regularization (const typename Architecture_t::Matrix_t &A, ERegularization R) -> decltype(Architecture_t::L1Regularization(A))
 Evaluate the regularization functional for a given weight matrix. More...