SVM(3)									 1								    SVM(3)

The SVM class

INTRODUCTION

CLASS SYNOPSIS

       SVM

       SVM

       Constants

       o const integer$SVM::C_SVC0

       o const integer$SVM::NU_SVC1

       o const integer$SVM::ONE_CLASS2

       o const integer$SVM::EPSILON_SVR3

       o const integer$SVM::NU_SVR4

       o const integer$SVM::KERNEL_LINEAR0

       o const integer$SVM::KERNEL_POLY1

       o const integer$SVM::KERNEL_RBF2

       o const integer$SVM::KERNEL_SIGMOID3

       o const integer$SVM::KERNEL_PRECOMPUTED4

       o const integer$SVM::OPT_TYPE101

       o const integer$SVM::OPT_KERNEL_TYPE102

       o const integer$SVM::OPT_DEGREE103

       o const integer$SVM::OPT_SHRINKING104

       o const integer$SVM::OPT_PROPABILITY105

       o const integer$SVM::OPT_GAMMA201

       o const integer$SVM::OPT_NU202

       o const integer$SVM::OPT_EPS203

       o const integer$SVM::OPT_P204

       o const integer$SVM::OPT_COEF_ZERO205

       o const integer$SVM::OPT_C206

       o const integer$SVM::OPT_CACHE_SIZE207

       Methods

       o public SVM::__construct (void	)

       o public float svm::crossvalidate (array  $problem, int	$number_of_folds)

       o public array SVM::getOptions (void  )

       o public bool SVM::setOptions (array  $params)

       o public SVMModel svm::train (array  $problem, [array  $weights])

PREDEFINED CONSTANTS

SVM CONSTANTS

	      o SVM::C_SVC -The basic C_SVC SVM type. The default, and a good starting point

	      o SVM::NU_SVC -The NU_SVC type uses a different, more flexible, error weighting

	      o SVM::ONE_CLASS -One class SVM type. Train just on a single class, using outliers as negative examples

	      o SVM::EPSILON_SVR -A SVM type for regression (predicting a value rather than just a class)

	      o SVM::NU_SVR -A NU style SVM regression type

	      o SVM::KERNEL_LINEAR -A very simple kernel, can work well on large document classification problems

	      o SVM::KERNEL_POLY -A polynomial kernel

	      o SVM::KERNEL_RBF -The common Gaussian RBD kernel. Handles non-linear problems well and is a good default for classification

	      o SVM::KERNEL_SIGMOID  -A  kernel  based on the sigmoid function. Using this makes the SVM very similar to a two layer sigmoid based
		neural network

	      o SVM::KERNEL_PRECOMPUTED -A precomputed kernel - currently unsupported.

	      o SVM::OPT_TYPE -The options key for the SVM type

	      o SVM::OPT_KERNEL_TYPE -The options key for the kernel type

	      o SVM::OPT_DEGREE -

	      o SVM::OPT_SHRINKING -Training parameter, boolean, for whether to use the shrinking heuristics

	      o SVM::OPT_PROBABILITY -Training parameter, boolean, for whether to collect and use probability estimates

	      o SVM::OPT_GAMMA -Algorithm parameter for Poly, RBF and Sigmoid kernel types.

	      o SVM::OPT_NU -The option key for the nu parameter, only used in the NU_ SVM types

	      o SVM::OPT_EPS -The option key for the Epsilon parameter, used in epsilon regression

	      o SVM::OPT_P -Training parameter used by Episilon SVR regression

	      o SVM::OPT_COEF_ZERO -Algorithm parameter for poly and sigmoid kernels

	      o SVM::OPT_C -The option for the cost parameter that controls tradeoff between errors and generality - effectively the  penalty  for
		misclassifying training examples.

	      o SVM::OPT_CACHE_SIZE -Memory cache size, in MB

PHP Documentation Group 														    SVM(3)