Frontiers to the learning of nonparametric hidden Markov models

Hidden Markov models (HMMs) are flexible tools for clustering dependent data coming from unknown populations, allowing nonparametric modelling of the population densities. Identifiability fails when data are in fact independent, and we study the frontier between learnable and unlearnable two-state nonparametric HMMs. Interesting new phenomena emerge when the cluster distributions are modelled via density functions (the `emission' densities) belonging to standard smoothness classes compared to the multinomial setting [1]. Notably, in contrast to the multinomial setting previously considered, the identification of a direction separating the two emission densities becomes a critical, and challenging, issue. Surprisingly, it is possible to "borrow strength" from estimators of the smoothest density to improve estimation of the other. We conduct precise analysis of minimax rates, showing a transition depending on compared smoothnesses of the emission densities.