define_neighborhood: Calculate neighborhood relations

In fcampelo/MOEADr: Component-Wise MOEA/D Implementation

Calculate neighborhood relations

Description

Calculates neighborhood relations for the MOEADr package

Usage

define_neighborhood(neighbors, v.matrix, iter)

Arguments

neighbors	List containing the decomposition method parameters. This list must contain the following key-value pairs: neighbors$name, type of neighborhood to use. The following types are currently available: neighbors$name = "lambda": defines the neighborhood using the distance matrix for the weight vectors. The calculation is performed only once for the entire run. neighbors$name = "x": defines the neighborhood using the distance matrix for the incumbent solution associated with each subproblem. In this case the calculation is performed at each iteration. neighbors$T: Neighborhood size. The value of neighbors$T must be smaller than the number of subproblems. neighbors$delta.p: Probability of sampling from the neighborhood when performing variation. Must be a scalar value between 0 and 1.
v.matrix	matrix of vectors to be used for defining the neighborhoods.
iter	iteration counter of the MOEA/D

Details

This routine calculates the neighborhood relations for the MOEA/D.

Warning: this routine may access (but not directly modify) variables from the calling environment.

Value

List containing the matrix of selection probabilities (P) and the matrix of neighborhoods (B).

References

F. Campelo, L.S. Batista, C. Aranha (2020): The MOEADr Package: A Component-Based Framework for Multiobjective Evolutionary Algorithms Based on Decomposition. Journal of Statistical Software doi: 10.18637/jss.v092.i06

fcampelo/MOEADr documentation built on Jan. 9, 2023, 6 a.m.