Desarrollo de un framework para evolución diferencial
Fecha
2021-09-10
Autores
Título de la revista
ISSN de la revista
Título del volumen
Editor
Jaén: Universidad de Jaén
Resumen
Los algoritmos de evolución diferencial son un reciente paradigma en el campo de las Metaheurísticas que se
caracterizan por su versatilidad en la resolución de problemas de optimización. En este trabajo se presenta un
estudio del estado del arte exhaustivo sobre este paradigma junto con el análisis, modelado y desarrollo de un
framework para evolución diferencial, denominado DEFramework. Es un sistema capaz de realizar
experimentación tanto con funciones test de optimización como con conjuntos de datos reales, permitiendo la
creación de nuevas variantes mediante la combinación de mútliples operadores en los algoritmos implementados.
Como resultado se ha obtenido un framework con utilidad real para la investigación en evolución diferencial.
Differential evolution algorithms are a recent paradigm in the field of Metaheuristics that are characterized by their versatility in solving optimization problems. This work presents an exhaustive state of the art study on this paradigm together with the analysis, modeling and development of a framework for differential evolution, called DEFramework. It is a system capable of conducting experimentation with both optimization test functions and real data sets, allowing the creation of new variants by combining multiple operators in the implemented algorithms. As a result, a framework with real utility for differential evolution research has been obtained.
Differential evolution algorithms are a recent paradigm in the field of Metaheuristics that are characterized by their versatility in solving optimization problems. This work presents an exhaustive state of the art study on this paradigm together with the analysis, modeling and development of a framework for differential evolution, called DEFramework. It is a system capable of conducting experimentation with both optimization test functions and real data sets, allowing the creation of new variants by combining multiple operators in the implemented algorithms. As a result, a framework with real utility for differential evolution research has been obtained.
Descripción
Palabras clave
Sistemas de la Información