Subspace estimation and prediction methods for hidden Markov models

TL;DR

This paper investigates subspace estimation techniques for hidden Markov models, focusing on their geometric structure and demonstrating the consistency of the proposed algorithms for estimating transition and emission probabilities.

Contribution

It introduces a subspace estimation method for HMMs, analyzing its geometric properties and proving the consistency of the transition and emission matrix estimates.

Findings

Transition and emission matrices are estimated consistently up to similarity.

The m-step linear predictor converges to the true predictor.

The geometric structure of HMMs influences the estimation process.

Abstract

Hidden Markov models (HMMs) are probabilistic functions of finite Markov chains, or, put in other words, state space models with finite state space. In this paper, we examine subspace estimation methods for HMMs whose output lies a finite set as well. In particular, we study the geometric structure arising from the nonminimality of the linear state space representation of HMMs, and consistency of a subspace algorithm arising from a certain factorization of the singular value decomposition of the estimated linear prediction matrix. For this algorithm, we show that the estimates of the transition and emission probability matrices are consistent up to a similarity transformation, and that the $m$-step linear predictor computed from the estimated system matrices is consistent, i.e., converges to the true optimal linear $m$-step predictor.