Sample-efficient neural likelihood-free Bayesian inference of implicit HMMs
This work addresses a specific bottleneck in Bayesian inference for implicit HMMs, offering a more efficient alternative for researchers in computational statistics and machine learning, though it is incremental in improving existing likelihood-free methods.
The authors tackled the problem of inaccurate posterior predictive distribution estimation in likelihood-free inference for Hidden Markov Models (HMMs) by proposing a novel method that directly learns the intractable posterior distribution of hidden states using an autoregressive-flow, achieving sample efficiency and quality comparable to a more computationally expensive SMC algorithm.
Likelihood-free inference methods based on neural conditional density estimation were shown to drastically reduce the simulation burden in comparison to classical methods such as ABC. When applied in the context of any latent variable model, such as a Hidden Markov model (HMM), these methods are designed to only estimate the parameters, rather than the joint distribution of the parameters and the hidden states. Naive application of these methods to a HMM, ignoring the inference of this joint posterior distribution, will thus produce an inaccurate estimate of the posterior predictive distribution, in turn hampering the assessment of goodness-of-fit. To rectify this problem, we propose a novel, sample-efficient likelihood-free method for estimating the high-dimensional hidden states of an implicit HMM. Our approach relies on learning directly the intractable posterior distribution of the hidden states, using an autoregressive-flow, by exploiting the Markov property. Upon evaluating our approach on some implicit HMMs, we found that the quality of the estimates retrieved using our method is comparable to what can be achieved using a much more computationally expensive SMC algorithm.