Analysis of Multiclass Support Vector Machines
Shigeo Abe · Institutional Repositories DataBase (IRDB) · 2002
Since support vector machines for pattern classification are based on two-class classification problems, unclassifiable regions exist when extended to problems with more than two classes.In our previous work, to solve this problem, we developed fuzzy support vector machines for one-against-all and pairwise classifications, introducing membership functions.In this paper, for one-against-all classification, we show that fuzzy support vector machines are equivalent to support vector machines with continuous decision functions.For pairwise classification, we discuss the relations between decision-tree-based support vector machines: DDAGs and ADAGs and compare classification performance of fuzzy support vector machines with that of ADAGs.