In this recipe, we will continue to use the telecom churn dataset as the input data source to train the support vector machine. For those who have not prepared the dataset, please refer to Chapter 7, Classification 1 - Tree, Lazy, and Probabilistic, for more details.