A new trigonometric kernel function for support vector machine

In the last few years, various types of machine learning algorithms, such as Support Vector Machine (SVM), Regression (SVR), and Non-negative Matrix Factorization (NMF) have been introduced. The kernel approach is an effective method for increasing classification accuracy algorithms. This paper introduces a family one-parameter functions improving SVM classification. proposed function consists trigonometric term differs from all existing functions. We show this positive definite function. Finally, we evaluate based on new kernel, Gaussian polynomial convex combination datasets. Empirical results that mixed achieve best accuracy. Moreover, some numerical performing SVR are presented.