In this example a two-class linear support vector machine classifier is trained
on randomly generated data. As training algorithm the OCAS solver is used with
the SVM regularization parameter C=10 and the bias term in the classification
rule switched off. The solver iterates until the relative duality gap falls
below epsilon=1e-3. The trained classifier is then used to compute outputs on
the training examples and the primal SVM objective function is computed.

For more details on the OCAS solver see
 V. Franc, S. Sonnenburg. Optimized Cutting Plane Algorithm for Large-Scale Risk
 Minimization.The Journal of Machine Learning Research, vol. 10,
 pp. 2157--2192. October 2009.
