SuperNN
0.7.0
|
Public Member Functions | |
Params () | |
Public Attributes | |
double | beta |
mu multiply/divide factor More... | |
Eigen::VectorXd | gradient |
Error gradient matrix. More... | |
Eigen::MatrixXd | hessian |
Quasi-Hessian matrix. More... | |
double | mu |
Current value of mu. More... | |
double | mu_max |
Maximum value of mu. More... | |
double | mu_min |
Minimum value of mu. More... | |
double | mu_zero |
Initial value for mu. More... | |
Eigen::VectorXd | t_jacob_line |
Transposed Jacobian line. More... | |
Eigen::VectorXd | weights |
Weight matrix (same as the network weights, but in a suitable format to the algorithm. More... | |
Eigen::VectorXd | weights_backup |
Backup of the weights, for backtracking. More... | |
Definition at line 385 of file training.cpp.
|
inline |
Definition at line 387 of file training.cpp.
double SuperNN::NBN::Params::beta |
mu multiply/divide factor
Definition at line 414 of file training.cpp.
Eigen::VectorXd SuperNN::NBN::Params::gradient |
Error gradient matrix.
Definition at line 398 of file training.cpp.
Eigen::MatrixXd SuperNN::NBN::Params::hessian |
Quasi-Hessian matrix.
Definition at line 392 of file training.cpp.
double SuperNN::NBN::Params::mu |
Current value of mu.
Definition at line 411 of file training.cpp.
double SuperNN::NBN::Params::mu_max |
Maximum value of mu.
Definition at line 420 of file training.cpp.
double SuperNN::NBN::Params::mu_min |
Minimum value of mu.
Definition at line 417 of file training.cpp.
double SuperNN::NBN::Params::mu_zero |
Initial value for mu.
Definition at line 408 of file training.cpp.
Eigen::VectorXd SuperNN::NBN::Params::t_jacob_line |
Transposed Jacobian line.
Definition at line 395 of file training.cpp.
Eigen::VectorXd SuperNN::NBN::Params::weights |
Weight matrix (same as the network weights, but in a suitable format to the algorithm.
Definition at line 402 of file training.cpp.
Eigen::VectorXd SuperNN::NBN::Params::weights_backup |
Backup of the weights, for backtracking.
Definition at line 405 of file training.cpp.