Search

This paper presents a grey particle swarm optimization learning algorithms to self classify different input patterns as correct subgroups. In our learning stratagem, a grey relational analysis (GRA) is combined with Particle swarm optimization (PSO) to develop a classification system. In our research, the GRA is more effective and accurate than the common-used Euclidean norm measure. The evolutional PSO is a very powerful learning algorithm to optimize the classifying task for complex, irregular and high dimensional input patterns. Several artificial data clustering examples compared with k-means method is presented to demonstrate the robustness of the proposed grey particle swarm optimization clustering algorithm. A real image classifying problem is also provided with the grey particle swarm optimization clustering algorithm. [ABSTRACT FROM AUTHOR]