A Bayesian Approach to Temporal Data Clustering using Hidden Markov Models

Cen Li and Gautam Biswas

International Conference on Machine Learning (ICML 2000), Stanford, California, Pages 543-550

Abstract - This paper presents clustering techniques that partition temporal data into homegeneous groups, and constructs state based profiles for each group in the Hidden Markov model (HMM) framework. We propose a Bayesian HMM clustering methodology that improves upon existing HMM clustering by incorporating HMM model size selection into clustering control structure to derive better cluster models and partition. Experimental results indicate the effectiveness of our methodology.