All Packages  Class Hierarchy  This Package  Previous  Next  Index  WEKA's home

Class weka.attributeSelection.GeneticSearch

java.lang.Object
   |
   +----weka.attributeSelection.ASSearch
           |
           +----weka.attributeSelection.GeneticSearch

public class GeneticSearch

extends ASSearch

implements StartSetHandler, OptionHandler

For more information see:

David E. Goldberg (1989). Genetic algorithms in search, optimization and machine learning. Addison-Wesley.

Valid options are:

-Z
Sets the size of the population. (default = 20).

-G
Sets the number of generations to perform. (default = 5).

-C
Sets the probability that crossover will occur. (default = .6).

-M
Sets the probability that a feature will be toggled on/off.

-R
Sets how frequently reports will be generated. Eg, setting the value to 5 will generate a report every 5th generation.

(default = number of generations).

-S
Sets the seed for random number generation.

Version:

$Revision: 1.6 $

Author:

Mark Hall (mhall@cs.waikato.ac.nz)

GeneticSearch()

Constructor.

getCrossoverProb()

get the probability of crossover

getMaxGenerations()

get the number of generations

getMutationProb()

get the probability of mutation

getOptions()

Gets the current settings of ReliefFAttributeEval.

getPopulationSize()

get the size of the population

getReportFrequency()

get how often repports are generated

getSeed()

get the value of the random number generator's seed

getStartSet()

Returns a list of attributes (and or attribute ranges) as a String

listOptions()

Returns an enumeration describing the available options

search(ASEvaluation, Instances)

Searches the attribute subset space using a genetic algorithm.

setCrossoverProb(double)

set the probability of crossover

setMaxGenerations(int)

set the number of generations to evaluate

setMutationProb(double)

set the probability of mutation

setOptions(String[])

Parses a given list of options.

setPopulationSize(int)

set the population size

setReportFrequency(int)

set how often reports are generated

setSeed(int)

set the seed for random number generation

setStartSet(String)

Sets a starting set of attributes for the search.

toString()

returns a description of the search

 public GeneticSearch()

Constructor. Make a new GeneticSearch object

 public Enumeration listOptions()

Returns an enumeration describing the available options

Returns:

an enumeration of all the available options

 public void setOptions(String options[]) throws Exception

Parses a given list of options. Valid options are:

-Z
Sets the size of the population. (default = 20).

-G
Sets the number of generations to perform. (default = 5).

-C
Sets the probability that crossover will occur. (default = .6).

-M
Sets the probability that a feature will be toggled on/off.

-R
Sets how frequently reports will be generated. Eg, setting the value to 5 will generate a report every 5th generation.

(default = number of generations).

-S
Sets the seed for random number generation.

Parameters:

options - the list of options as an array of strings

Throws: Exception

if an option is not supported

 public String[] getOptions()

Gets the current settings of ReliefFAttributeEval.

Returns:

an array of strings suitable for passing to setOptions()

 public void setStartSet(String startSet) throws Exception

Sets a starting set of attributes for the search. It is the search method's responsibility to report this start set (if any) in its toString() method.

Parameters:

startSet - a string containing a list of attributes (and or ranges), eg. 1,2,6,10-15.

Throws: Exception

if start set can't be set.

 public String getStartSet()

Returns a list of attributes (and or attribute ranges) as a String

Returns:

a list of attributes (and or attribute ranges)

 public void setSeed(int s)

set the seed for random number generation

Parameters:

s - seed value

 public int getSeed()

get the value of the random number generator's seed

Returns:

the seed for random number generation

 public void setReportFrequency(int f)

set how often reports are generated

Parameters:

f - generate reports every f generations

 public int getReportFrequency()

get how often repports are generated

Returns:

how often reports are generated

 public void setMutationProb(double m)

set the probability of mutation

Parameters:

m - the probability for mutation occuring

 public double getMutationProb()

get the probability of mutation

Returns:

the probability of mutation occuring

 public void setCrossoverProb(double c)

set the probability of crossover

Parameters:

c - the probability that two population members will exchange genetic material

 public double getCrossoverProb()

get the probability of crossover

Returns:

the probability of crossover

 public void setMaxGenerations(int m)

set the number of generations to evaluate

Parameters:

m - the number of generations

 public int getMaxGenerations()

get the number of generations

Returns:

the maximum number of generations

 public void setPopulationSize(int p)

set the population size

Parameters:

p - the size of the population

 public int getPopulationSize()

get the size of the population

Returns:

the population size

 public String toString()

returns a description of the search

Returns:

a description of the search as a String

Overrides:

toString in class Object

 public int[] search(ASEvaluation ASEval,
                     Instances data) throws Exception

Searches the attribute subset space using a genetic algorithm.

Parameters:

ASEvaluator - the attribute evaluator to guide the search

data - the training instances.

Returns:

an array (not necessarily ordered) of selected attribute indexes

Throws: Exception

if the search can't be completed

Overrides:

search in class ASSearch

All Packages  Class Hierarchy  This Package  Previous  Next  Index  WEKA's home