Optimum-Path Forest: A Novel and Powerful Framework for Supervised Graph-based Pattern Recognition Techniques

João Paulo PapaAlexandre X. Falcão

We present here a novel framework for graph-based pattern recognition techniques called Optimum-Path Forest (OPF), which has been demonstrated to be superior than traditional supervised pattern recognition techniques, such as Artificial Neural Networks using Multilayer Perceptrons and Support Vector Machines, in terms of both accuracy and execution times. The OPF-based classifiers model the problem of the pattern recognition as a computation of an optimum-path forest in a graph induced by the dataset samples, achieving very good results in complex situations, i.e., in which we have a large amount of overlapped regions. Results in several real and synthetic datasets show the robustness of the OPF-based classifiers against the above ones.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: