Inteligência de enxame aplicada a problemas complexos de análise de dados: agrupamento multiobjetivo e mineração de regras de associação em data streams

• Autor(a) principal: Cunha, Danilo Souza da
• Data de Publicação: 2019
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações do Mackenzie
• Texto Completo: http://dspace.mackenzie.br/handle/10899/24292
• topic: inteligência de enxame; colônias de abelhas; colônias de bactérias; agrupamento multiobjetivo; regras de associação; fluxos de dados; CNPQ::CIENCIAS EXATAS E DA TERRA; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO

The increasing accumulation of data, from the most diverse sources, brings the need to develop new and robust data analysis tools. Among the many data analysis problems, two of them have been receiving special attention over the past years, both in academic and in business: multi-objective data clustering; and mining of data streams. At the same time, bio-inspired al-gorithms have been successfully applied in solving complex problems, such as those previously mentioned. It is in this context that this thesis proposes two new versions of swarm intelligence algorithms to solve these problems. More specifically, an algorithm inspired by the collective behavior of bees is used to tackle the multi-objective clustering problem, while a tool inspired by the collective decision-making of bacteria is used to mine association rules in data streams. The algorithms are potentially applicable to the cited current problems and the main thesis' contribution is the investigation of algorithms' adaptation and subsequent application to the already mentioned data mining complex problems. This document provides the theoretical basis necessary for the development and understanding of the research, introduces the algorithms, presents the results obtained, the conclusion of the performance achieved and brings a reflection of the future steps to be taken.