Adaptation in symbolic dynamic systems for pattern classification

This paper addresses the problem of pattern classification in the symbolic dynamic domain, where the patterns of interest are represented by probabilistic finite state automata (PFSA) with possibly dissimilar algebraic structures. A combination of Dirichlet and multinomial distributions is used to model the uncertainties due to the finite length approximation of symbol strings. The classifier algorithm follows the structure of a Bayes model and has been validated on a simulation test bed.