Universal, nonlinear, mean-square prediction of Markov processes

摘要

We estimate the best, nonlinear, mean-square predictor for aMarkov process from an observed, finite realization of the process whenthe true Markov order is unknown. In particular, we propose an universalminimum complexity estimator, which does not know the true Markov order,and yet delivers the same statistical performance as that delivered by aminimum complexity estimator, which knows the true Markov order