Algoritmos exatos para o problema de edição de p-Clusters

• Ano de defesa: 2015
• Autor(a) principal: Cabral, Lucidio dos Anjos Formiga
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Informática; Programa de Pós-Graduação em Informática; UFPB
• Palavras-chave em Português: Edição de Grafos; Graph Editing; Clusterização; Branch-and-Cut; Branch-and-Price; Clustering; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/tede/7870

This work deals with the p-Cluster Editing Problem (p-CEP), which consists in performing a minimum number of editions (additions or removals of edges) in a graph G in order to transform it in a disjoint union of p cliques (clusters), where G and p are input data. p-CEP is a NP-Hard problem with applications in areas such as computational biology and machine learning. To solve this problem, we propose two new mathematical formulations and improvements in a formulation from the literature, as well as new valid inequalities. The three formulations were studied both theoretically, by comparing their linear relaxations, and practically, by implementing three exact algorithms: two based on Branch-and-Cut (BC) and one based on Branch-and-Price (BP). The proposed algorithms were tested in instances with up to 211 vertices. The results show the performance of the algorithms according to the graph density and the ratio between p and the number of vertices. Overall, the BC algorithms were superior to the BP algorithm. However, the latter obtained the best dual bounds in some instances.