On Computation with Higher-order Markov Chains

Abstract

Categorical data sequences occur in many real world applications. The major problem in using higher-order Markov chain model is that the number of parameters increases exponentially with respect to the order of the model. In this paper, we propose a higher-order Markov chain model for modeling categorical data sequences where the number of model parameters increases linearly with respect to the order of the model. We present efficient estimation methods based on linear programming for the model parameters. The model is then compared with other existing models with simulated sequences and DNA data sequences of mouse.