M. Malyutov, P. Grosu, and T. Zhang

SCOT stationary distribution evaluation for some examples

We call Stochastic COntext Trees (abbreviated as SCOT) n-Markov Chains with every state of a string independent of the symbols in its more remote past than the context of length determined by the preceding symbols of this state. Previous somewhat confusing names for SCOT were VLMC, PST, CTW. We estimated SCOT parameters for testing homogeneity of data strings in No. 4, vol. 13 of this journal. Our more efficient SCOT fitting algorithm will be exposed elsewhere. Here, we postulate SCOT models and study their convergence without fitting SCOT from data sets. A SCOT stationary distribution over contexts is iteratively evaluated here explicitly in several examples. Our main tool is a 1-MC generated by the SCOT with the set of contexts as its state space.

КЛЮЧЕВЫЕ СЛОВА: Variable Memory Length Markov Chain, Stochastic Context Trees, Stationary Distribution of Contexts