{
"title": "FIR Filter Design via Linear Complementarity Problem, Messy Genetic Algorithm, and Ising Messy Genetic Algorithm",
"authors": "A.M. Al-Fahed Nuseirat, R. Abu-Zitar",
"country": null,
"institution": null,
"volume": "9",
"journal": "International Journal of Electrical, Computer, Energetic, Electronic and Communication Engineering",
"pagesStart": 1430,
"pagesEnd": 1440,
"ISSN": "1307-6892",
"URL": "http:\/\/waset.org\/publications\/3072",
"abstract": "In this paper the design of maximally flat linear phase\r\nfinite impulse response (FIR) filters is considered. The problem is\r\nhandled with totally two different approaches. The first one is\r\ncompletely deterministic numerical approach where the problem is\r\nformulated as a Linear Complementarity Problem (LCP). The other\r\none is based on a combination of Markov Random Fields (MRF's)\r\napproach with messy genetic algorithm (MGA). Markov Random\r\nFields (MRFs) are a class of probabilistic models that have been\r\napplied for many years to the analysis of visual patterns or textures.\r\nOur objective is to establish MRFs as an interesting approach to\r\nmodeling messy genetic algorithms. We establish a theoretical result\r\nthat every genetic algorithm problem can be characterized in terms of\r\na MRF model. This allows us to construct an explicit probabilistic\r\nmodel of the MGA fitness function and introduce the Ising MGA.\r\nExperimentations done with Ising MGA are less costly than those\r\ndone with standard MGA since much less computations are involved.\r\nThe least computations of all is for the LCP. Results of the LCP,\r\nrandom search, random seeded search, MGA, and Ising MGA are\r\ndiscussed.",
"references": null,
"publisher": "World Academy of Science, Engineering and Technology",
"index": "International Science Index 9, 2007"
}