Density Ratio Hidden Markov Models
This addresses the problem of improving sequential classification for domains like speech recognition and bioinformatics, though it is incremental as it builds on existing HMM frameworks.
The paper tackled the classification performance drawback of Hidden Markov Models (HMMs) by applying density ratio estimation to bypass likelihood learning, resulting in a striking increase in discriminative performance while retaining probabilistic qualities.
Hidden Markov models and their variants are the predominant sequential classification method in such domains as speech recognition, bioinformatics and natural language processing. Being generative rather than discriminative models, however, their classification performance is a drawback. In this paper we apply ideas from the field of density ratio estimation to bypass the difficult step of learning likelihood functions in HMMs. By reformulating inference and model fitting in terms of density ratios and applying a fast kernel-based estimation method, we show that it is possible to obtain a striking increase in discriminative performance while retaining the probabilistic qualities of the HMM. We demonstrate experimentally that this formulation makes more efficient use of training data than alternative approaches.