A Sparse Modeling Approach to Based on Relevance Vect

In this paper, we compare two powerful kernel-based learning machines, support vector machines (SVM) and relevance vector machines (RVM), wi th in the framework of hidden Markov model-based speech recognit ion. Both machines provide nonl inear discriminative classification ability: the SVM by kernelbased margin maximization and the RVM using a Bayesian… CONTINUE READING