LIBIRWLS
|
Functions to train a semi parametric SVM using the IRWLS algorithm. More...
#include "IOStructures.h"
Go to the source code of this file.
Functions | |
int * | randomCentroids (svm_dataset dataset, properties props) |
Random selection of centroids for the semiparametric model. More... | |
int * | SGMA (svm_dataset dataset, properties props) |
Sparse Greedy Matrix Approximation algorithm. More... | |
double * | IRWLSpar (svm_dataset dataset, int *indexes, properties props) |
Iterative Re-Weighted Least Squares Algorithm. More... | |
void | printPSIRWLSInstructions (void) |
Print Instructions. More... | |
properties | parseTrainParameters (int *argc, char ***argv) |
It parses input command line to extract the parameters of the PSIRWLS algorithm. More... | |
model | calculatePSIRWLSModel (properties props, svm_dataset dataset, int *centroids, double *beta) |
It converts the result of the PSIRWLS algorithm into a model struct. More... | |
Functions to train a semi parametric SVM using the IRWLS algorithm.
model calculatePSIRWLSModel | ( | properties | props, |
svm_dataset | dataset, | ||
int * | centroids, | ||
double * | beta | ||
) |
It converts the result of the PSIRWLS algorithm into a model struct.
It converts the result into a model struct.
props | The training parameters. |
dataset | The training set. |
centroids | of the selected centroids by the SGMA algorithm. |
beta | The weights of every centroid obtained with the IRWLS algorithm. |
double* IRWLSpar | ( | svm_dataset | dataset, |
int * | indexes, | ||
properties | props | ||
) |
Iterative Re-Weighted Least Squares Algorithm.
IRWLS procedure to obtain the weights of the semi parametric model.
dataset | The training set. |
indexes | The indexes of the centroids selected by the SGMA algorithm. |
props | The struct with the training parameters. |
properties parseTrainParameters | ( | int * | argc, |
char *** | argv | ||
) |
It parses input command line to extract the parameters of the PSIRWLS algorithm.
It parses input command line to extract the parameters.
argc | The number of words of the command line. |
argv | The list of words of the command line. |
void printPSIRWLSInstructions | ( | void | ) |
Print Instructions.
It shows PSIRWLS-train command line instructions in the standard output.
int* randomCentroids | ( | svm_dataset | dataset, |
properties | props | ||
) |
Random selection of centroids for the semiparametric model.
It creates a random permutation and selects the first elements to be the indexes of the centroids of the semiparametric model.
dataset | The training set. |
props | The struct with the training parameters. |
int* SGMA | ( | svm_dataset | dataset, |
properties | props | ||
) |
Sparse Greedy Matrix Approximation algorithm.
Sparse Greedy Matrix Approximation algorithm to select the basis elements of the semi parametric model.
dataset | The training set. |
props | The struct with the training parameters. |