X-GAT: uma ferramenta baseada em XML para otimização com algoritmos genéticos.

• Ano de defesa: 2011
• Autor(a) principal: Soares Junior, Amilcar
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraí­ba; BR; Informática; Programa de Pós-Graduação em Informática; UFPB
• Palavras-chave em Português: Engenharia de software; Inteligência artificial bio-inspirada; Otimização; Software engineering; Bio-inspired artificial intelligence; Optimization; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/tede/6155

In many different fields of science we are faced to optimization problems. Many approaches have been proposed to solve such types of problems, including the use of Evolutionary Algorithms (AEs) and Genetic Algorithms (GAs). However, just a low quantity of works tried to create generic tools, capable to be reused in many different problems. Most of the current proposals show different implementations of GAs or AEs applied to specific purposes. Raising all these assertions, it is verified the need for the development of optimization tools that are capable of solving any kind of optimization problems. In such context, the main objective of this work is the development of an optimization tool, named X-GAT (XML based Genetic Algorithm Toolkit), that is: capable of solving any kind of optimization problem; configurable; operating system portable; extensible; a framework that helps the implementation of AEs; and can be used to build heterogenic systems. Using some computational techniques, parts of the algorithm can be abstracted, preventing that much time is spent on coding and validating the optimization technique. Aiming to achieve such objectives, some tools and techniques were used: Java programming language and its reflection API; the data description language XML (eXtensible Markup Language); and software design patterns. In order to verify and validate that the developed tool attended the proposed objectives, many different optimization problems were chosen. First, it is shown the optimization of some mathematical functions that have known optimum value. Then, the X-GAT tool is applied to the calibration of for rainfall-runoff models input parameters, which are common problems in hydrology. Finally, the X-GAT tool is used in an optimization process of input parameters of clustering algorithms for grouping trajectories points of moving objects. The motivation behind applying these algorithms is the addition of semantic information to spatiotemporal data. From the results obtained in many different fields of science, the proposed toolkit showed to be flexible and robust, in addition to being able to be easily applied in many problems, once it is properly configured.