parsimony_Crossover: Crossover operators in GA-PARSIMONY

Description Usage Arguments Value Author(s) See Also

Description

Functions implementing particular crossover genetic operator for GA-PARSIMONY. Method uses for model parameters Heuristic Blending and random swapping for binary selected features.

Usage

1
parsimony_crossover(object, parents, alpha=0.1, perc_to_swap=0.5, ...)

Arguments

object

An object of class "ga_parsimony", usually resulting from a call to function ga_parsimony.

parents

A two-rows matrix of values indexing the parents from the current population.

alpha

A tuning parameter for the Heuristic Blending outer bounds [Michalewicz, 1991]. Typical and default value is 0.1.

perc_to_swap

Percentage of features for swapping in the crossovering process.

...

Further arguments passed to or from other methods.

Value

Return a list with two elements:

children

Matrix of dimension 2 times the number of decision variables containing the generated offsprings;

fitnessval

Vector of length 2 containing the fitness validation values for the offsprings. A value NA is returned if an offspring is different (which is usually the case) from the two parents.

fitnesstst

Vector of length 2 containing the fitness with the test database (if it was supplied), for the offsprings. A value NA is returned if an offspring is different (which is usually the case) from the two parents.

complexity

Vector of length 2 containing the model complexity for the offsprings. A value NA is returned if an offspring is different (which is usually the case) from the two parents.

Author(s)

Francisco Javier Martinez de Pison. fjmartin@unirioja.es. EDMANS Group. http://www.mineriadatos.com

See Also

ga_parsimony


GAparsimony documentation built on Dec. 3, 2019, 5:07 p.m.