Optimize Selection (Evolutionary)

From Rapid-I-Wiki

A genetic algorithm for feature selection.

Contents 1 Description 2 Input 3 Output 4 Parameters

Description

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

* Generate an initial population consisting of population_size individuals. Each attribute is switched on with probability p_initialize
• 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.
• 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.
• 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.

Input

• example set in: expects: ExampleSetMetaData: #examples: = 0; #attributes: 0
• attribute weights in: optional: AttributeWeights
• through 1:

Output

• example set out:
• weights:
• performance:

Parameters

• use exact number of attributes:
  Determines if only combinations containing this numbers of attributes should be tested.
  Range: boolean; default: false
• restrict maximum:
  If checked the maximal number of attributes might be restricted. Otherwise all combinations of all number of attributes are generated and tested.
  Range: boolean; default: false
• min number of attributes:
  Determines the minimum number of features used for the combinations.
  Range: integer; 1-+?; default: 1
• max number of attributes:
  Determines the maximum number of features used for the combinations.
  Range: integer; 1-+?; default: 1
• exact number of attributes:
  Determines the exact number of features used for the combinations.
  Range: integer; 1-+?; default: 1
• initialize with input weights:
  Indicates if this operator should look for attribute weights in the given input and use the input weights of all known attributes as starting point for the optimization.
  Range: boolean; default: false
• population size:
  Number of individuals per generation.
  Range: integer; 1-+?; default: 5
• maximum number of generations:
  Number of generations after which to terminate the algorithm.
  Range: integer; 1-+?; default: 30
• use early stopping:
  Enables early stopping. If unchecked, always the maximum number of generations is performed.
  Range: boolean; default: false
• generations without improval:
  Stop criterion: Stop after n generations without improval of the performance.
  Range: integer; 1-+?; default: 2
• normalize weights:
  Indicates if the final weights should be normalized.
  Range: boolean; default: true
• use local random seed:
  Indicates if a local random seed should be used.
  Range: boolean; default: false
• local random seed:
  Specifies the local random seed
  Range: integer; 1-+?; default: 1992
• show stop dialog:
  Determines if a dialog with a button should be displayed which stops the run: the best individual is returned.
  Range: boolean; default: false
• user result individual selection:
  Determines if the user wants to select the final result individual from the last population.
  Range: boolean; default: false
• show population plotter:
  Determines if the current population should be displayed in performance space.
  Range: boolean; default: false
• plot generations:
  Update the population plotter in these generations.
  Range: integer; 1-+?; default: 10
• constraint draw range:
  Determines if the draw range of the population plotter should be constrained between 0 and 1.
  Range: boolean; default: false
• draw dominated points:
  Determines if only points which are not Pareto dominated should be painted.
  Range: boolean; default: true
• population criteria data file:
  The path to the file in which the criteria data of the final population should be saved.
  Range: filename
• maximal fitness:
  The optimization will stop if the fitness reaches the defined maximum.
  Range: real; 0.0-+?
• selection scheme:
  The selection scheme of this EA.
  Range: uniform, cut, roulette wheel, stochastic universal sampling, Boltzmann, rank, tournament, non dominated sorting; default: tournament
• tournament size:
  The fraction of the current population which should be used as tournament members.
  Range: real; 0.0-1.0
• start temperature:
  The scaling temperature .
  Range: real; 0.0-+?
• dynamic selection pressure:
  If set to true the selection pressure is increased to maximum during the complete optimization run.
  Range: boolean; default: true
• keep best individual:
  If set to true, the best individual of each generations is guaranteed to be selected for the next generation (elitist selection).
  Range: boolean; default: false
• save intermediate weights:
  Determines if the intermediate best results should be saved.
  Range: boolean; default: false
• intermediate weights generations:
  Determines if the intermediate best results should be saved. Will be performed every k generations for a specified value of k.
  Range: integer; 1-+?; default: 10
• intermediate weights file:
  The file into which the intermediate weights will be saved.
  Range: filename
• p initialize:
  Initial probability for an attribute to be switched on.
  Range: real; 0.0-1.0
• p mutation:
  Probability for an attribute to be changed (-1: 1 / numberOfAtt).
  Range: real; -1.0-1.0
• p crossover:
  Probability for an individual to be selected for crossover.
  Range: real; 0.0-1.0
• crossover type:
  Type of the crossover.
  Range: one_point, uniform, shuffle; default: uniform