MCMC-Interactive Variational Inference

Abstract

Leveraging well-established MCMC strategies, we propose MCMC-interactive variational inference (MIVI) to not only estimate the posterior in a time constrained manner, but also facilitate the design of MCMC transitions. Constructing a variational distribution followed by a short Markov chain that has parameters to learn, MIVI takes advantage of the complementary properties of variational inference and MCMC to encourage mutual improvement. On one hand, with the variational distribution locating high posterior density regions, the Markov chain is optimized within the variational inference framework to efficiently target the posterior despite a small number of transitions. On the other hand, the optimized Markov chain with considerable flexibility guides the variational distribution towards the posterior and alleviates its underestimation of uncertainty. Furthermore, we prove the optimized Markov chain in MIVI admits extrapolation, which means its marginal distribution gets closer to the true posterior as the chain grows. Therefore, the Markov chain can be used separately as an efficient MCMC scheme. Experiments show that MIVI not only accurately and efficiently approximates the posteriors but also facilitates designs of stochastic gradient MCMC and Gibbs sampling transitions.

arXiv comment25 pages, 7 figures, 3 tables

References in corpus (23)

• Adam: A Method for Stochastic Optimization
• Auto-Encoding Variational Bayes
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Discussion of "Least angle regression" by Efron et al
• Least Angle Regression
• Rejoinder to "Least angle regression" by Efron et al
• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• Markov Chain Monte Carlo and Variational Inference: Bridging the Gap
• Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks
• Preconditioned Stochastic Gradient Langevin Dynamics for Deep Neural Networks
• Lognormal and Gamma Mixed Negative Binomial Regression
• Auto-Encoding Sequential Monte Carlo
• Continuous-Time Flows for Efficient Inference and Density Estimation
• Approximate Inference with Amortised MCMC
• A Contrastive Divergence for Combining Variational Inference and MCMC
• Learning Model Reparametrizations: Implicit Variational Inference by Fitting MCMC distributions
• Variationally Inferred Sampling Through a Refined Bound for Probabilistic Programs