com.rapidminer.operator.learner.functions
Class LogisticRegressionOptimization

java.lang.Object
  extended by com.rapidminer.tools.math.optimization.ec.es.ESOptimization
      extended by com.rapidminer.operator.learner.functions.LogisticRegressionOptimization
All Implemented Interfaces:
Optimization

public class LogisticRegressionOptimization
extends ESOptimization

Evolutionary Strategy approach for optimization of the logistic regression problem.

Author:
Ingo Mierswa, Tobias Malbrecht

Field Summary
 
Fields inherited from class com.rapidminer.tools.math.optimization.ec.es.ESOptimization
BOLTZMANN_SELECTION, CUT_SELECTION, GAUSSIAN_MUTATION, INIT_TYPE_MAX, INIT_TYPE_MIN, INIT_TYPE_ONE, INIT_TYPE_RANDOM, INIT_TYPE_ZERO, MUTATION_TYPES, NO_MUTATION, NON_DOMINATED_SORTING_SELECTION, PARAMETER_CROSSOVER_PROB, PARAMETER_GENERATIONS_WITHOUT_IMPROVAL, PARAMETER_KEEP_BEST, PARAMETER_MAX_GENERATIONS, PARAMETER_MUTATION_TYPE, PARAMETER_POPULATION_SIZE, PARAMETER_SELECTION_TYPE, PARAMETER_SHOW_CONVERGENCE_PLOT, PARAMETER_SPECIFIY_POPULATION_SIZE, PARAMETER_TOURNAMENT_FRACTION, PARAMETER_USE_EARLY_STOPPING, POPULATION_INIT_TYPES, RANK_SELECTION, ROULETTE_WHEEL, SELECTION_TYPES, SPARSITY_MUTATION, STOCHASTIC_UNIVERSAL, SWITCHING_MUTATION, TOURNAMENT_SELECTION, UNIFORM_SELECTION
 
Constructor Summary
LogisticRegressionOptimization(ExampleSet exampleSet, boolean addIntercept, int initType, int maxIterations, int generationsWithoutImprovement, int popSize, int selectionType, double tournamentFraction, boolean keepBest, int mutationType, double crossoverProb, boolean showConvergencePlot, RandomGenerator random, LoggingHandler logging)
          Creates a new evolutionary optimization.
 
Method Summary
 PerformanceVector evaluateIndividual(Individual individual)
          Subclasses must implement this method to calculate the fitness of the given individual.
 PerformanceVector getPerformance()
           
 void nextIteration()
          This method is invoked after each evaluation.
 LogisticRegressionModel train()
           
 
Methods inherited from class com.rapidminer.tools.math.optimization.ec.es.ESOptimization
evaluate, evaluate, evaluateAll, getBestFitnessEver, getBestFitnessInGeneration, getBestPerformanceEver, getBestValuesEver, getGeneration, getMax, getMin, getParameterTypes, getPopulation, getValueType, increaseCurrentEvaluationCounter, increaseTotalEvaluationCounter, optimize, setMax, setMin, setValueType
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

LogisticRegressionOptimization

public LogisticRegressionOptimization(ExampleSet exampleSet,
                                      boolean addIntercept,
                                      int initType,
                                      int maxIterations,
                                      int generationsWithoutImprovement,
                                      int popSize,
                                      int selectionType,
                                      double tournamentFraction,
                                      boolean keepBest,
                                      int mutationType,
                                      double crossoverProb,
                                      boolean showConvergencePlot,
                                      RandomGenerator random,
                                      LoggingHandler logging)
Creates a new evolutionary optimization.

Method Detail

evaluateIndividual

public PerformanceVector evaluateIndividual(Individual individual)
Description copied from class: ESOptimization
Subclasses must implement this method to calculate the fitness of the given individual. Please note that null might be returned for non-valid individuals. The fitness will be maximized.

Specified by:
evaluateIndividual in class ESOptimization

train

public LogisticRegressionModel train()
                              throws OperatorException
Throws:
OperatorException

nextIteration

public void nextIteration()
Description copied from class: ESOptimization
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 ESOptimization

getPerformance

public PerformanceVector getPerformance()


Copyright © 2001-2009 by Rapid-I