com.rapidminer.operator.learner.functions.kernel.evosvm
Class PSOSVMOptimization

java.lang.Object
  com.rapidminer.tools.math.optimization.ec.pso.PSOOptimization
      com.rapidminer.operator.learner.functions.kernel.evosvm.PSOSVMOptimization

All Implemented Interfaces:

Optimization

public class PSOSVMOptimization
extends PSOOptimization

PSO approach for SVM optimization. Currently only classification problems are supported.

Author:

Ingo Mierswa Exp $

Constructor Summary

PSOSVMOptimization(ExampleSet exampleSet, Kernel kernel, double c, int maxIterations, int generationsWithoutImprovement, int popSize, double inertiaWeight, double localWeight, double globalWeight, boolean dynamicInertiaWeight, boolean showPlot, RandomGenerator random)
Creates a new evolutionary SVM optimization.

Method Summary

PerformanceVector	evaluateIndividual(double[] individual) Evaluates the individuals of the given population.
EvoSVMModel	getModel(double[] alphas) Returns a model containing all support vectors, i.e. the examples with non-zerp alphas.
void	nextIteration() This method is invoked after each evaluation.

Methods inherited from class com.rapidminer.tools.math.optimization.ec.pso.PSOOptimization

createInitialPopulation, getBestFitnessEver, getBestFitnessInGeneration, getBestPerformanceEver, getBestValuesEver, getGeneration, optimize, reinit, setMaxValue, setMinValue

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

PSOSVMOptimization

public PSOSVMOptimization(ExampleSet exampleSet,
                          Kernel kernel,
                          double c,
                          int maxIterations,
                          int generationsWithoutImprovement,
                          int popSize,
                          double inertiaWeight,
                          double localWeight,
                          double globalWeight,
                          boolean dynamicInertiaWeight,
                          boolean showPlot,
                          RandomGenerator random)

Creates a new evolutionary SVM optimization.

Method Detail

nextIteration

public void nextIteration()

Description copied from class: PSOOptimization

This method is invoked after each evaluation. The default implementation does nothing but subclasses might implement this method to support online plotting or logging.

Overrides:

nextIteration in class PSOOptimization

evaluateIndividual

public PerformanceVector evaluateIndividual(double[] individual)

Evaluates the individuals of the given population.

Specified by:

evaluateIndividual in class PSOOptimization

getModel

public EvoSVMModel getModel(double[] alphas)

Returns a model containing all support vectors, i.e. the examples with non-zerp alphas.