SVMAttributeEval

Package

weka.attributeSelection

Synopsis

SVMAttributeEval :

Evaluates the worth of an attribute by using an SVM classifier. Attributes are ranked by the square of the weight assigned by the SVM. Attribute selection for multiclass problems is handled by ranking attributes for each class seperately using a one-vs-all method and then "dealing" from the top of each pile to give a final ranking.

For more information see:

I. Guyon, J. Weston, S. Barnhill, V. Vapnik (2002). Gene selection for cancer classification using support vector machines. Machine Learning. 46:389-422.

Available in Weka 3.6.x - 3.7.1. Available via the package management system for Weka >= 3.7.2 (SVMAttributeEval)

Options

The table below describes the options available for SVMAttributeEval.

Option

Description

attsToEliminatePerIteration

Constant rate of attribute elimination.

complexityParameter

C complexity parameter to pass to the SVM

epsilonParameter

P epsilon parameter to pass to the SVM

filterType

filtering used by the SVM

percentThreshold

Threshold below which percent elimination reverts to constant elimination.

percentToEliminatePerIteration

Percent rate of attribute elimination.

toleranceParameter

T tolerance parameter to pass to the SVM

Capabilities

The table below describes the capabilites of SVMAttributeEval.

Capability

Supported

Class

Nominal class, Binary class, Missing class values

Attributes

Numeric attributes, Unary attributes, Binary attributes, Missing values, Empty nominal attributes

Min # of instances

1