OpenANN
1.1.0
An open source library for artificial neural networks.
|
Creates several types of matrices for compression. More...
#include <CompressionMatrixFactory.h>
Public Types | |
enum | Transformation { DCT, GAUSSIAN, SPARSE_RANDOM, AVERAGE, EDGE } |
Public Member Functions | |
CompressionMatrixFactory () | |
CompressionMatrixFactory (int inputDim, int paramDim, Transformation transformation=DCT) | |
void | createCompressionMatrix (Eigen::MatrixXd &cm) |
Public Attributes | |
bool | compress |
enum OpenANN::CompressionMatrixFactory::Transformation | transformation |
int | inputDim |
int | paramDim |
Creates several types of matrices for compression.
A compression matrix can either be used to compress a data vector or the parameters of a Learner.
OpenANN::CompressionMatrixFactory::CompressionMatrixFactory | ( | ) |
OpenANN::CompressionMatrixFactory::CompressionMatrixFactory | ( | int | inputDim, |
int | paramDim, | ||
Transformation | transformation = DCT |
||
) |
void OpenANN::CompressionMatrixFactory::createCompressionMatrix | ( | Eigen::MatrixXd & | cm) |
bool OpenANN::CompressionMatrixFactory::compress |
int OpenANN::CompressionMatrixFactory::inputDim |
int OpenANN::CompressionMatrixFactory::paramDim |
enum OpenANN::CompressionMatrixFactory::Transformation OpenANN::CompressionMatrixFactory::transformation |