Goal programming approaches to support vector machines

Support vector machines (SVMs) are gaining much popularity as effective methods in machine learning. In pattern classification problems with two class sets, their basic idea is to find a maximal margin separating hyperplane which gives the greatest separation between the classes in a high dimensional feature space. However, the idea of maximal margin separation is not quite new: in 1960's the multi-surface method (MSM) was suggested by Mangasarian. In 1980's, linear classifiers using goal programming were developed extensively. This paper considers SVMs from a viewpoint of goal programming, and proposes a new method based on the total margin instead of the shortest distance between learning data and separating hyperplane.