Support Vector Machines are non-probabilistic discriminative classification models. The main concept is to find boundaries in the feature space that separate samples of different classes with a maximum margin. The boundaries are modeled by so called support vectors, where this classification model gets its name from.


Online Resources