1 Support Vector Machine Solvers

The support vector machine (SVM) algorithm (Cortes and Vapnik, 1995) is probably the most widely used kernel learning algorithm. It achieves relatively robust pattern recognition performance using well-established concepts in optimization theory. Despite this mathematical classicism, the implementation of efficient SVM solvers has diverged from the classical methods of numerical optimization. This divergence is common to virtually all learning algorithms. The numerical optimization literature focuses on the asymptotic performance: how quickly the accuracy of the solution increases with computing time. In the case of learning algorithms, two other factors mitigate the impact of optimization accuracy.