Rule extraction from Support Vector Machines : a geometric approach

This paper presents a new approach to rule extraction from Support Vector Machines. SVMs have been applied successfully in many areas with excellent generalization results; rule extraction can offer explanation capability to SVMs. We propose to approximate the SVM classification boundary through querying followed by clustering, searching and then to extract rules by solving an optimization problem. Theoretical proof and experimental results then indicate that the rules can be used to validate the SVM results, since maximum fidelity with high accuracy can be achieved.