Resumo

FAJARDO CALDERIN, Jenny et al. Deceptive Dynamic Optimization Problems, experimentation with Metaheuristics. Rev cuba cienc informat [online]. 2015, vol.9, suppl.1, pp. 98-109. ISSN 2227-1899.

In the optimization field there are a number of problems called NP, which are those that can be solved by a nondeterministic polynomial time algorithm for a resolution. Because the real world is not static, but dynamic, the need to bring these problems to test reality is created, hence arise dynamic optimization problems (PODs). One of the classic optimization problems that exist, are disappointing or deceptive problems, which are problems of test generation from binary XOR. They are called deceptive because the algorithms have a hard time getting improvements, because when solution is improved heuristic evaluation worsens the objective function. In recent years there has been an increasing interest in the modeling of dynamic optimization problems and their solution with metaheuristic algorithms. Therefore, the objective of this research is to analyze the behavior of the classical metaheuristics disappointing compared to dynamic problems, specifically with five disappointing functions and evaluating the performance of algorithms using non-parametric statistical test. Furthermore a comparison of the results of the best algorithm with two the state of art algorithms are usually done to solve dynamic optimization: Adaptive Hill Climbing Memetic Algorithm y Self Organized Random Immigrants Genetic Algorithm.

Palavras-chave : dynamic optimization problems; deceptive; metaheuristics algorithm.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )