Simpler knowledge-based support vector machines

Le, Quoc; Smola, Alexander; Gaertner, Thomas

Simpler knowledge-based support vector machines

Date

2006

Authors

Le, Quoc

Smola, Alexander

Gaertner, Thomas

Publisher

Association for Computing Machinery Inc (ACM)

Abstract

If appropriately used, prior knowledge can significantly improve the predictive accuracy of learning algorithms or reduce the amount of training data needed. In this paper we introduce a simple method to incorporate prior knowledge in support vector machines by modifying the hypothesis space rather than the optimization problem. The optimization problem is amenable to solution by the constrained concave convex procedure, which finds a local optimum. The paper discusses different kinds of prior knowledge and demonstrates the applicability of the approach in some characteristic experiments.

Keywords

Keywords: Constraint theory; Data acquisition; Knowledge acquisition; Learning algorithms; Optimization; Problem solving; Constrained concave convex procedure; Optimization problem; Predictive accuracy; Support vector machines