In this paper, we propose a new method to choose the effective samples for support vector machines (SVM) training based on regression tree in audio classification task. The objective is to reduce the training time of SVM by choosing effective examples from the training set and to balance the number of training points of binary classes. One obvious advantage of such method is that it provides a flexible framework to implement the choice procedure based on the training data for a given classification task. We test the performances of our new method on a dataset composed of about 6-hour audio data which illustrate that the computation time can be significantly reduced without a significant decrease in the prediction accuracy.