The Support vector machine (SVM) is a supervised learning algorithm used to separate groups of data with a margin which is as clear as possible to ensure it is more likely to generalize well to examples it has never seen before.
As seen in figure 1 there are three different margins to achieve separation of the two groups of data. The brown margin provides a better separation between the two groups of data compared to the pink or the yellow margin, as it maximizes the orthogonal distance to the closest points in each classified group. The SVM model is able to find a margin similar to the brown margin, and is able to give a clearer separation of data.
- 1. Marsland S. Machine Learning. Hoboken: CRC Press; 2011 ISBN:1420067184 9781420067187