An improved adaptive Support Vector Machine algorithm with combinational fuzzy C-means clustering

In order to improve the training efficiency to the data set, an improved adaptive Support Vector Machine (SVM) algorithm with combinational Fuzzy C-means Clustering is proposed. With multi-layer fuzzy C-means clustering algorithm original data are pretreated to remove the training data, which has no contribution to the classification. The remaining data are used to complete the training work for SVM to obtain the optimal hyper-plane. Besides, the parameter adaptive optimization algorithm has both increased the flexibility of parameter selection for SVM and enhanced the convergence speed. In the end, derived from the comparison of testing performance using the data set from the database of Statlog, the experiment result indicates that the proposed algorithm can both shorten the training time and provides high accuracy and excellent generalization, also it can keep the distribution of original data set at the same time.