#include <rf_full.hpp>
Inheritance diagram for RFFull:
Public Member Functions | |
RFFull () | |
Default constructor for the RF. | |
RFFull (Types::Input c) | |
A constructor taking only a center point for the RF. | |
RFFull (Types::Input c, const boost::program_options::variables_map &vm) | |
A constructor taking a center point and a variables_map. | |
RFFull (const RFFull &, Types::Input c, const boost::program_options::variables_map &vm) | |
a copy constructor, sort of. | |
virtual Types::RValue | getActivation (Types::Input x) const |
Return the activation for a test point. | |
virtual Types::RValue | learn (Types::Input x, Types::Output y, Types::RValueP w) |
Update the receptive field and have local model update itself. | |
Types::Matrix | getAlpha () const |
stupid get functions for copy constructor. | |
Types::Matrix | getD () const |
Types::Matrix | getM () const |
Types::Input | getCenter () const |
Private Attributes | |
Types::Matrix | dwdM |
Types::Matrix | dJ2dM |
Types::Matrix | dJdM |
Types::Matrix | dJ2_2dM_2 |
Types::Matrix | aux |
Types::Matrix | dw_2dM_2 |
Types::Matrix | dJ_2dM_2 |
Meta Learning Sufficient Statistics | |
This group of variables control the behavior of the algorithm.
All paramters are defined in the paper (see PAPER). | |
Types::Matrix | alpha |
Learning rate for the distance matrix updates. | |
Types::Matrix | meta_h |
Meta Update Params. | |
Types::Matrix | b |
RF Descriptors | |
These variables describe the receptive field.
They include the center, distance matrix, and local regression model | |
Types::Matrix | D |
the distance matrix of the RF | |
Types::Matrix | M |
The Cholesky decomposition of D. | |
Friends | |
std::ostream & | operator<< (std::ostream &out, const RFFull &rf) |
print the center and distance matrix on a line |
This class handles updates of the distance matrix for the receptive field and hold the local model associated with the field.
Definition at line 39 of file rf_full.hpp.
RFFull::RFFull | ( | Types::Input | c | ) |
RFFull::RFFull | ( | const RFFull & | , | |
Types::Input | c, | |||
const boost::program_options::variables_map & | vm | |||
) |
virtual Types::RValue RFFull::getActivation | ( | Types::Input | x | ) | const [inline, virtual] |
Return the activation for a test point.
This currently uses ublas vectro expressions, but should be changed to use blas calls once everything is working correctly.
x | the input vector |
Implements RF.
Definition at line 121 of file rf_full.hpp.
References RF::center, M, and RF::temp2.
Types::RValue RFFull::learn | ( | Types::Input | x, | |
Types::Output | y, | |||
Types::RValueP | w | |||
) | [virtual] |
Update the receptive field and have local model update itself.
Update the distance matrix as in Equation 3.6 and Table 4
x | the input vector | |
y | the actual output |
Implements RF.
Definition at line 128 of file rf_full.cpp.
References RF::a_E, RF::a_G, RF::a_H, RF::a_pk, alpha, aux, b, RF::center, Model::check_derivatives(), D, dJ2_2dM_2, dJ2dM, dJ_2dM_2, dJdM, dw_2dM_2, dwdM, RF::e_2, RF::gamma, Model::get_a_zz(), Model::get_e(), Model::get_e_cv(), Model::get_z(), RF::H_temp, RF::lambda, Model::learn(), RF::localModel, M, RF::meta, RF::meta_alpha, meta_h, RF::q, RF::temp2, Model::updateNumProjections(), RF::W, and RF::z.
Here is the call graph for this function:
Types::Matrix RFFull::M [private] |
The Cholesky decomposition of D.
This is used for updating the distance matrix.
Definition at line 80 of file rf_full.hpp.
Referenced by getActivation(), getM(), learn(), and RFFull().