Solving Imbalance Data Classification Problem by Particle Swarm Optimization Support Vector Machine

A database has a plenty of hidden knowledge, which can be used in decision making to support commerce, research and other activities. Classification analysis performs a very important rule in pattern recognition field as one core research topics. Algorithms like support vector machine (SVM) and artificial network (ANN) have been proposed to perform binary classification according to the distribution. But these traditional classification algorithms can hardly performs the satisfied result for imbalanced dataset. In this paper, we proposed to perform a model on the basis of Particle Swarm Optimization (PSO) and support vector machine (SVM) for a large imbalanced dataset. This model is named PSOSVC (Particle Swarm Optimization support vector classification) model. Recently, PSO is proposed used as a meta heuristic frame work for the large imbalanced classification. The SVM also shows high performance in balanced binary classification, so a novel model combined both support vector classification (SVC) and PSO is introduced to improve the classification accuracy. In this paper, G-mean is used to evaluate the final result. Performance in the final part of this paper the proposed method is compared with some conventional models, the results will show the high performance for imbalanced dataset classification by using the proposed method.