E-mail Spam Filtering Using Support Vector Machines with Selection of Kernel Function Parameters

Support Vector Machines (SVM) is a powerful classification technique in data mining and has been successfully applied to many real-world applications. Parameter selection of SVM will affect classification performance much during training process. However, parameter selection of SVM is usually identified by experience or grid search (GS). GS is simple and easily implemented, but it is very time-consuming. In this study, Taguchi method is proposed for improving GS and used to optimize the SVMbased E-mail Spam Filtering model. It is easy to implement by orthogonal arrays without iteration. A real-world mail dataset is selected to demonstrate the effectiveness and feasibility of the method. The results show that the Taguchi method can find the effective model with high classification accuracy and good robustness.