(PECL svm >= 0.1.0) Create a SVMModel based on training data
(PECL svm >= 0.1.0) Set training parameters
(PECL svm >= 0.1.0) Construct a new SVM object
(PECL svm >= 0.1.0) Return the current training parameters
(PECL svm >= 0.1.0) Test training params on subsets of the training data.