An RPCL-based approach for Markov model identification with unknown state number

This paper presents an alternative identification approach for the Markov m odel studied in [3]. Our approach estimates the state sequence and model pa rameters with the help of a clustering analysis by the rival penalized comp etitive learning (RPCL) algorithm [4]. Compared to the method in [3], this new approach not only extends the model from scalar states to multi-dimensi onal ones, but also makes the model identification with the correct number of states decided automatically. The experiments have shown that it works w ell.