Support Vector Machine is a classification method in supervised learning that seeks to use support vectors (cases close to the boundary) to find an optimal hyperplane separating items from different classes.