Online modeling based on support vector machine

Abstract

Support vector machine (SVM) is a new method based on statistical learning theory. Online algorithms for training SVM are efficient to run, easy to implement comparing with batch algorithms. Presently online algorithms usually do not provide with the ability to explicitly control the number of support vectors. A modified online algorithm for SVM is proposed, witch has a budget parameter to explicitly control the number of support vectors. The proposed algorithm was applied to construct intelligent model of helicopter. It is shown by simulation that the modified online algorithm can reduce the number of support vectors effectively with similar generalization ability.