A hierarchical method for multi-class support vector machines

We introduce a framework, which we call Divide-by-2 (DB2), for extending support vector machines (SVM) to multi-class problems. DB2 offers an alternative to the standard one-against-one and one-against-rest algorithms. For an N class problem, DB2 produces an N - 1 node binary decision tree where nodes represent decision boundaries formed by N -1 SVM binary classifiers. This tree structure allows us to present a generalization and a time complexity analysis of DB2. Our analysis and related experiments show that, DB2 is faster than one-againstone and one-against-rest algorithms in terms of testing time, significantly faster than oneagainst-rest in terms of training time, and that the cross-validation accuracy of DB2 is comparable to these two methods.

• Type: Conference
• Year: 2004
• Where: ICML
• Authors: Volkan Vural, Jennifer G. Dy
• Keywords: Binary Decision Tree, ICML 2004, Machine Learning, Nodes Represent Decision, SVM Binary Classifiers