Memetic Algorithm with Hungarian Matching Based Crossover and Diversity Preservation

Ruiz, Emmanuel Romero; Segura, Carlos

doi:10.13053/cys-22-2-2951

Cited by 6 publications

(3 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the first variant [34], a multi-objective approach based on a tuple that consists of fitness and diversity was applied and the selection process chooses survivors at random from among the non-dominated individuals. Note that subsequently, a simpler strategy based only on fitness also behaved properly [31], so the step of distinguishing between penalized and non-penalized individuals seems to be key to proper performance. In the case of GP, both the accuracy and sizes of trees are important.…”

Section: Proposalmentioning

confidence: 99%

GP-DMD: a genetic programming variant with dynamic management of diversity

Nieto-Fuentes

Segura

2022

Genet Program Evolvable Mach

Self Cite

View full text Add to dashboard Cite

The proper management of diversity is essential to the success of Evolutionary Algorithms. Specifically, methods that explicitly relate the amount of diversity maintained in the population to the stopping criterion and elapsed period of execution, with the aim of attaining a gradual shift from exploration to exploitation, have been particularly successful.However, in the area of Genetic Programming (GP), the performance of this design principle has not been studied. In this paper, a novel GP variant, Genetic Programming with Dynamic Management of Diversity (GP-DMD), is presented. GP-DMD applies this novel design principle through a replacement strategy that combines penalties based on distance-like functions with a multi-objective Pareto selection based on maximizing accuracy and minimizing tree size. The experimental validation based on the symbolic regression task considers several state-of-the-art approaches and shows clear improvements both in terms of accuracy and simplicity. The effects of GP-DMD on the dynamics of the population are also analyzed, revealing the reasons for the superiority of GP-DMD. As in other fields of Evolutionary Computation, this design principle contributes significantly to the area of GP.

show abstract

Section: Proposalmentioning

confidence: 99%

GP-DMD: a genetic programming variant with dynamic management of diversity

Nieto-Fuentes

Segura

2022

Genet Program Evolvable Mach

Self Cite

View full text Add to dashboard Cite

show abstract

“…The second variant of our optimizer includes a replacement phase that takes diversity into account. Particularly, the replacement phase is similar to the one devised for the Graph Partitioning Problem [ 10 ]. This replacement operator (Algorithm 8) operates as follows.…”

Section: Population-based Proposalsmentioning

confidence: 99%

“…Regarding the diversity management, two strategies that involve different ways of controlling the diversity are applied. The first one, which was successfully devised for the Graph Partitioning Problem [ 10 ], enforces a large contribution to diversity for every member of the population, whereas the second one, which is a novel proposal, considers the creation of clusters, meaning that some close members are maintained in the population. The lower entropy induced by the formation of clusters [ 11 ] allows the promotion of a larger degree of intensification, which is a key to success.…”

Section: Introductionmentioning

confidence: 99%

Metaheuristics in the Optimization of Cryptographic Boolean Functions

López-López

Sosa-Gómez

Segura

et al. 2020

Entropy

Self Cite

View full text Add to dashboard Cite

Generating Boolean Functions (BFs) with high nonlinearity is a complex task that is usually addresses through algebraic constructions. Metaheuristics have also been applied extensively to this task. However, metaheuristics have not been able to attain so good results as the algebraic techniques. This paper proposes a novel diversity-aware metaheuristic that is able to excel. This proposal includes the design of a novel cost function that combines several information from the Walsh Hadamard Transform (WHT) and a replacement strategy that promotes a gradual change from exploration to exploitation as well as the formation of clusters of solutions with the aim of allowing intensification steps at each iteration. The combination of a high entropy in the population and a lower entropy inside clusters allows a proper balance between exploration and exploitation. This is the first memetic algorithm that is able to generate 10-variable BFs of similar quality than algebraic methods. Experimental results and comparisons provide evidence of the high performance of the proposed optimization mechanism for the generation of high quality BFs.

show abstract

A parallel memetic algorithm with explicit management of diversity for the job shop scheduling problem

Constantino

Segura

2021

Appl Intell

View full text Add to dashboard Cite

Memetic Algorithm with Hungarian Matching Based Crossover and Diversity Preservation

Cited by 6 publications

References 22 publications

GP-DMD: a genetic programming variant with dynamic management of diversity

GP-DMD: a genetic programming variant with dynamic management of diversity

Metaheuristics in the Optimization of Cryptographic Boolean Functions

A parallel memetic algorithm with explicit management of diversity for the job shop scheduling problem

Contact Info

Product

Resources

About