Made available in DSpace on 2016-03-15T19:37:34Z (GMT). No. of bitstreams: 1 Ana Karina Fontes Prior.pdf: 415415 bytes, checksum: a6ecb97b982ab886cc421abdc943c8ac (MD5) Previous issue date: 2010-12-22 / Fundo Mackenzie de Pesquisa / The large number of data generated by people and organizations has stimulated the research on effective and automatic methods of knowledge extraction from databases. This dissertation proposes two new bioinspired techniques, named cPSC and oPSC, based on the Particle Swarm Optimization Algorithm (PSO) to solve data clustering problems. The proposed algorithms are applied to data and text clustering problems and their performances are compared with a standard algorithm from the literature. The results allow us to conclude that the proposed algorithms are competitive with those already available in literature, but bring benefits such as automatic determination of the number of groups on the dataset and a search for the best partitioning of the dataset considering an explicit cost function. / A grande quantidade de dados gerados por pessoas e organizações tem estimulado a pesquisa sobre métodos efetivos e automáticos de extração de conhecimentos a partir de bases de dados. Essa dissertação propõe duas novas técnicas bioinspiradas, denominadas cPSC e oPSC, baseadas no algoritmo de otimização por enxame de partículas (PSO - Particle Swarm Optimization) para resolver problemas de agrupamento de dados. Os algoritmos propostos são aplicados a problemas de agrupamento de dados e textos, e seus desempenhos são comparados com outros propostos na literatura específica. Os resultados obtidos nos permitem concluir que os algoritmos propostos são competitivos com aqueles já disponíveis na literatura, porém trazem outros benefícios como a determinação automática do número de grupos nas bases e a efetuação de uma busca pelo melhor particionamento possível da base considerando uma função de custo explícita.

The Master Thesis deals with logistic module innovation of information system ERP. The principle of innovation is based on implementation of heuristic algorithms which solve Travel Salesman Problems (TSP). The software MATLAB is used for analysis and tests of these algorithms. The goal of Master Thesis is the comparison of selections algorithm, which are suitable for economic purposes (accuracy of solution, speed of calculation and memory demands).