Information Accessibility and Cryptic Processes: Linear Combinations of Causal States

Abstract

We show in detail how to determine the time-reversed representation of a stationary hidden stochastic process from linear combinations of its forward-time $\epsilon$-machine causal states. This also gives a check for the $k$-cryptic expansion recently introduced to explore the temporal range over which internal state information is spread.

6 pages, 9 figures, 2 tables; http://users.cse.ucdavis.edu/~cmg/compmech/pubs/iacplcocs.htm