Support vector machines

From DrugPedia: A Wikipedia for Drug discovery

Support Vector Machine (SVM) is a supervised learning algorithm developed by Vladimir Vapnik and his co-workers at AT&T Bell Labs in the mid 90's. Since their inception, they have continuously been shown to outperform many prior learning algorithms in both classification, and regression applications. In fact, the elegance and the rigorous mathematical foundations from optimization and statistical learning theory have propelled SVMs to the very forefront of the machine learning field within the last decade.

At their core, SVMs are a method for creating a predictor function from a set of training data where the function itself can be a binary, a multi-category, or even a general regression predictor. To accomplish this mathematical feat, SVMs find a hypersurface (for example, a plane in 2D) which attempts to split the positive and negative examples with the largest possible margin on all sides of the (hyper)plane.