Validación de un método heurístico de optimización basado en un sistema de infección por virus
Date
Authors
Subject
virus infection
heuristic optimization method
bacteriophage
Genetic Algorithms
Deceptive Functions
infección por virus
método heurístico de optimización
bacteriófago
algoritmos genético
funciones engañosas
heuristic optimization method
bacteriophage
Genetic Algorithms
Deceptive Functions
infección por virus
método heurístico de optimización
bacteriófago
algoritmos genético
funciones engañosas
Language:
Journal Title
Journal ISSN
Volume Title
Publisher
Instituto Tecnológico de Santo Domingo (INTEC)
En los problemas de optimización se pueden utilizar diferentes algoritmos para encontrar la solución adecuada. Pero cuando se trata de problemas de complejidad media y alta, como los NP difíciles, las técnicas conocidas como los Algoritmos Genéticos (GA) tiene dificultad en converger o llegar a una solución. En este artículo se presenta la implementación de un Sistema de Virus (VS, por sus siglas en inglés Virus System), que se desarrolla con un nuevo enfoque para resolver problemas de optimización simulando la forma en que un organismo es atacado por un virus. La analogía del VS se aplica a dos tipos de problemas, Onemax, de diferentes longitudes de bits y funciones engañosas (Deceptive Functions) con el objetivo de comprobar su funcionamiento y su potencia de convergencia. Este método es comparado con un GA inspirado en el crecimiento de corales marinos. El VS ha logrado conseguir resultados de alta precisión, con una convergencia del 100 % en ambos problemas y con considerables mejoras comparados con los obtenidos con el GA.
In optimization problems, different algorithms can be used to find the right solution. But when it comes to problems of medium and high complexity, such as difficult NPs, techniques known as Genetic Algorithms (GA) have a hard time converging or coming up with a solution. This article presents the implementation of a Virus System (VS, for its acronym in English Virus System), which is developed with a new approach to solve optimization problems simulating the way an organism is attacked by a virus. The VS analogy is applied to two types of problems, Onemax, of different bit lengths and deceptive functions (Deceptive Functions) with the aim of checking their operation and their convergence power. This method is compared to a GA inspired by the growth of marine corals. The VS has managed to achieve high precision results, with 100% convergence in both problems and with considerable improvements compared to those obtained with the GA.
In optimization problems, different algorithms can be used to find the right solution. But when it comes to problems of medium and high complexity, such as difficult NPs, techniques known as Genetic Algorithms (GA) have a hard time converging or coming up with a solution. This article presents the implementation of a Virus System (VS, for its acronym in English Virus System), which is developed with a new approach to solve optimization problems simulating the way an organism is attacked by a virus. The VS analogy is applied to two types of problems, Onemax, of different bit lengths and deceptive functions (Deceptive Functions) with the aim of checking their operation and their convergence power. This method is compared to a GA inspired by the growth of marine corals. The VS has managed to achieve high precision results, with 100% convergence in both problems and with considerable improvements compared to those obtained with the GA.
Description
Type
info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
Nota
info:eu-repo/semantics/publishedVersion
Nota
Source
Science, Engineering and Applications; Vol 3 No 1 (2020): Science, Engineering and Applications; 85-112
Ciencia, Ingenierías y Aplicaciones; Vol. 3 Núm. 1 (2020): Ciencia, Ingenierías y Aplicaciones; 85-112
2636-2171
2636-218X
10.22206/cyap.2020.v3i1
Ciencia, Ingenierías y Aplicaciones; Vol. 3 Núm. 1 (2020): Ciencia, Ingenierías y Aplicaciones; 85-112
2636-2171
2636-218X
10.22206/cyap.2020.v3i1