maswchains

MA-SW-Chains — Memetic Algorithm with persistent Solis-Wets local search chains

maswchains is a memetic algorithm based on MA-SW-Chains. Each individual maintains its own persistent Solis-Wets state across generations. After each DE/rand/1/bin phase, local search budget is distributed across individuals weighted by fitness rank and historical improvement.

Usage

maswchains(
  NP = 40,
  G = 100,
  data,
  class.name,
  F = 0.5,
  CR = 0.9,
  structure = c("nb", "tancl", "tan", "hc"),
  edgelist = NULL,
  ls_evals_ratio = 0.5,
  verbose = 25,
  ...
)

Parameters

NP · integer · default: "40"

Population size.

G · integer · default: "100"

Maximum generations.

data · data.frame · required

Training data frame.

class.name · character · required

Class variable column name.

F · numeric · default: "0.5"

Mutation factor for DE/rand/1. Range [0, 2].

CR · numeric · default: "0.9"

Crossover rate. Range [0, 1].

structure · character · default: "nb"

BN structure: "nb", "tan", "tancl", "hc".

edgelist · matrix

Optional custom BN topology (edges × 2 matrix).

ls_evals_ratio · numeric · default: "0.5"

Fraction of NP used as total local search budget per generation.

verbose · integer · default: "25"

Print progress every this many generations.

...

Extra arguments for bnclassify::tan_cl or bnclassify::tan_hc.

Return value

Object of class DE: Best, BestCLL, pobFinal, CLLPobFinal, N.evals, convergence, evaluations.

Example

library(dplbnDE)
data(car)

result <- maswchains(
  NP = 20, G = 25,
  data = car,
  class.name = names(car)[7],
  verbose = 5
)
print(result)