com.rapidminer.operator.features.selection
Class GeneticAlgorithm

java.lang.Object
  com.rapidminer.operator.Operator
      com.rapidminer.operator.OperatorChain
          com.rapidminer.operator.features.FeatureOperator
              com.rapidminer.operator.features.selection.AbstractGeneticAlgorithm
                  com.rapidminer.operator.features.selection.GeneticAlgorithm

All Implemented Interfaces:

ConfigurationListener, PreviewListener, ParameterHandler, LoggingHandler

public class GeneticAlgorithm
extends AbstractGeneticAlgorithm

A genetic algorithm for feature selection (mutation=switch features on and off, crossover=interchange used features). Selection is done by roulette wheel. Genetic algorithms are general purpose optimization / search algorithms that are suitable in case of no or little problem knowledge.
A genetic algorithm works as follows

1. Generate an initial population consisting of population_size individuals. Each attribute is switched on with probability p_initialize
2. For all individuals in the population
   • Perform mutation, i.e. set used attributes to unused with probability p_mutation and vice versa.
   • Choose two individuals from the population and perform crossover with probability p_crossover. The type of crossover can be selected by crossover_type.
3. Perform selection, map all individuals to sections on a roulette wheel whose size is proportional to the individual's fitness and draw population_size individuals at random according to their probability.
4. As long as the fitness improves, go to 2

If the example set contains value series attributes with blocknumbers, the whole block will be switched on and off.

Author:

Ingo Mierswa, Simon Fischer

Field Summary

static java.lang.String	PARAMETER_CROSSOVER_TYPE The parameter name for "Type of the crossover.
static java.lang.String	PARAMETER_EXACT_NUMBER_OF_ATTRIBUTES
static java.lang.String	PARAMETER_INITIALIZE_WITH_INPUT_WEIGHTS
static java.lang.String	PARAMETER_MAX_NUMBER_OF_ATTRIBUTES
static java.lang.String	PARAMETER_MIN_NUMBER_OF_ATTRIBUTES
static java.lang.String	PARAMETER_P_CROSSOVER The parameter name for "Probability for an individual to be selected for crossover.
static java.lang.String	PARAMETER_P_INITIALIZE The parameter name for "Initial probability for an attribute to be switched on.
static java.lang.String	PARAMETER_P_MUTATION The parameter name for "Probability for an attribute to be changed (-1: 1 / numberOfAtt).

Fields inherited from class com.rapidminer.operator.features.selection.AbstractGeneticAlgorithm

BOLTZMANN_SELECTION, CUT_SELECTION, NON_DOMINATED_SORTING_SELECTION, PARAMETER_DYNAMIC_SELECTION_PRESSURE, PARAMETER_GENERATIONS_WITHOUT_IMPROVAL, PARAMETER_INTERMEDIATE_WEIGHTS_FILE, PARAMETER_INTERMEDIATE_WEIGHTS_GENERATIONS, PARAMETER_KEEP_BEST_INDIVIDUAL, PARAMETER_MAXIMUM_NUMBER_OF_GENERATIONS, PARAMETER_POPULATION_SIZE, PARAMETER_SAVE_INTERMEDIATE_WEIGHTS, PARAMETER_SELECTION_SCHEME, PARAMETER_START_TEMPERATURE, PARAMETER_TOURNAMENT_SIZE, RANK_SELECTION, ROULETTE_WHEEL, SELECTION_SCHEMES, STOCHASTIC_UNIVERSAL, TOURNAMENT_SELECTION, UNIFORM_SELECTION

Fields inherited from class com.rapidminer.operator.features.FeatureOperator

PARAMETER_CONSTRAINT_DRAW_RANGE, PARAMETER_DRAW_DOMINATED_POINTS, PARAMETER_LOCAL_RANDOM_SEED, PARAMETER_MAXIMAL_FITNESS, PARAMETER_NORMALIZE_WEIGHTS, PARAMETER_PLOT_GENERATIONS, PARAMETER_POPULATION_CRITERIA_DATA_FILE, PARAMETER_SHOW_POPULATION_PLOTTER, PARAMETER_SHOW_STOP_DIALOG, PARAMETER_USER_RESULT_INDIVIDUAL_SELECTION

Constructor Summary

GeneticAlgorithm(OperatorDescription description)

Method Summary

Population	createInitialPopulation(ExampleSet es) Sets up a population of given size and creates ExampleSets with randomly selected attributes (the probability to be switched on is controlled by pInitialize).
protected PopulationOperator	getCrossoverPopulationOperator(ExampleSet eSet) Returns an operator that performs crossover.
protected PopulationOperator	getMutationPopulationOperator(ExampleSet eSet) Returns an operator that performs the mutation.
java.util.List<ParameterType>	getParameterTypes() Returns a list of ParameterTypes describing the parameters of this operator.

Methods inherited from class com.rapidminer.operator.features.selection.AbstractGeneticAlgorithm

getPostEvaluationPopulationOperators, getPostProcessingPopulationOperators, getPreEvaluationPopulationOperators, getPreProcessingPopulationOperators, solutionGoodEnough

Methods inherited from class com.rapidminer.operator.features.FeatureOperator

apply, createCleanClone, evaluate, getCheckForMaximum, getInnerOperatorCondition, getInputClasses, getMaxNumberOfInnerOperators, getMinNumberOfInnerOperators, getOutputClasses, getPopulation, getPopulationEvaluator, getRandom, setCheckForMaximum

Methods inherited from class com.rapidminer.operator.OperatorChain

addAddListener, addOperator, addOperator, checkDeprecations, checkIO, checkNumberOfInnerOperators, checkProperties, clearErrorList, cloneOperator, createExperimentTree, createProcessTree, getAllInnerOperators, getIndexOfOperator, getInnerOperatorForName, getInnerOperatorsXML, getNumberOfAllOperators, getNumberOfOperators, getOperator, getOperatorFromAll, getOperators, performAdditionalChecks, processFinished, processStarts, registerOperator, removeAddListener, removeOperator, shouldAddNonConsumedInput, shouldReturnInnerOutput, unregisterOperator

Methods inherited from class com.rapidminer.operator.Operator

addError, addValue, addWarning, apply, checkForStop, createExperimentTree, createFromXML, createMarkedExperimentTree, createMarkedProcessTree, createProcessTree, getAddOnlyAdditionalOutput, getApplyCount, getDeliveredOutputClasses, getDeprecationInfo, getDesiredInputClasses, getEncoding, getErrorList, getExperiment, getInput, getInput, getInput, getInputDescription, getIOContainerForInApplyLoopBreakpoint, getIODescription, getLog, getName, getOperatorClassName, getOperatorDescription, getParameter, getParameterAsBoolean, getParameterAsColor, getParameterAsDouble, getParameterAsFile, getParameterAsFile, getParameterAsInputStream, getParameterAsInt, getParameterAsMatrix, getParameterAsString, getParameterList, getParameters, getParameterType, getParent, getProcess, getStartTime, getStatus, getUserDescription, getValue, getValues, getXML, hasBreakpoint, hasBreakpoint, hasInput, inApplyLoop, isDebugMode, isEnabled, isExpanded, isParallel, isParameterSet, log, logError, logNote, logWarning, register, remove, rename, resume, setApplyCount, setBreakpoint, setEnabled, setExpanded, setInput, setListParameter, setOperatorParameters, setParameter, setParameters, setParent, setUserDescription, toString, writeXML

Methods inherited from class java.lang.Object

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

Field Detail

PARAMETER_P_INITIALIZE

public static final java.lang.String PARAMETER_P_INITIALIZE

The parameter name for "Initial probability for an attribute to be switched on."

See Also:

Constant Field Values

PARAMETER_P_MUTATION

public static final java.lang.String PARAMETER_P_MUTATION

The parameter name for "Probability for an attribute to be changed (-1: 1 / numberOfAtt)."

See Also:

Constant Field Values

PARAMETER_P_CROSSOVER

public static final java.lang.String PARAMETER_P_CROSSOVER

The parameter name for "Probability for an individual to be selected for crossover."

See Also:

Constant Field Values

PARAMETER_CROSSOVER_TYPE

public static final java.lang.String PARAMETER_CROSSOVER_TYPE

The parameter name for "Type of the crossover."

See Also:

Constant Field Values

PARAMETER_MAX_NUMBER_OF_ATTRIBUTES

public static final java.lang.String PARAMETER_MAX_NUMBER_OF_ATTRIBUTES

See Also:

Constant Field Values

PARAMETER_MIN_NUMBER_OF_ATTRIBUTES

public static final java.lang.String PARAMETER_MIN_NUMBER_OF_ATTRIBUTES

See Also:

Constant Field Values

PARAMETER_EXACT_NUMBER_OF_ATTRIBUTES

public static final java.lang.String PARAMETER_EXACT_NUMBER_OF_ATTRIBUTES

See Also:

Constant Field Values

PARAMETER_INITIALIZE_WITH_INPUT_WEIGHTS

public static final java.lang.String PARAMETER_INITIALIZE_WITH_INPUT_WEIGHTS

See Also:

Constant Field Values

Constructor Detail

GeneticAlgorithm

public GeneticAlgorithm(OperatorDescription description)

Method Detail

createInitialPopulation

public Population createInitialPopulation(ExampleSet es)
                                   throws OperatorException

Sets up a population of given size and creates ExampleSets with randomly selected attributes (the probability to be switched on is controlled by pInitialize).

Specified by:

createInitialPopulation in class FeatureOperator

Throws:

OperatorException

getMutationPopulationOperator

protected PopulationOperator getMutationPopulationOperator(ExampleSet eSet)
                                                    throws UndefinedParameterError

Returns an operator that performs the mutation. Can be overridden by subclasses.

Specified by:

getMutationPopulationOperator in class AbstractGeneticAlgorithm

Throws:

UndefinedParameterError

getCrossoverPopulationOperator

protected PopulationOperator getCrossoverPopulationOperator(ExampleSet eSet)
                                                     throws UndefinedParameterError

Returns an operator that performs crossover. Can be overridden by subclasses.

Specified by:

getCrossoverPopulationOperator in class AbstractGeneticAlgorithm

Throws:

UndefinedParameterError

getParameterTypes

public java.util.List<ParameterType> getParameterTypes()

Description copied from class: Operator

Returns a list of ParameterTypes describing the parameters of this operator. The default implementation returns an empty list if no input objects can be retained and special parameters for those input objects which can be prevented from being consumed.

Specified by:

getParameterTypes in interface ParameterHandler

Overrides:

getParameterTypes in class AbstractGeneticAlgorithm