Automated Variational Inference in Probabilistic Programming

Abstract

We present a new algorithm for approximate inference in probabilistic programs, based on a stochastic gradient for variational programs. This method is efficient without restrictions on the probabilistic program; it is particularly practical for distributions which are not analytically tractable, including highly structured distributions that arise in probabilistic programs. We show how to automatically derive mean-field probabilistic programs and optimize them, and demonstrate that our perspective improves inference efficiency over other algorithms.

References in corpus (5)

• Church: a language for generative models
• Stochastic Variational Inference
• Hybrid Variational/Gibbs Collapsed Inference in Topic Models
• Optimization of Structured Mean Field Objectives
• Variational Program Inference

Cited by in corpus (40)

• Variational Inference: A Review for Statisticians
• An Introduction to Variational Autoencoders
• Variational Inference with Normalizing Flows
• Automatic Differentiation Variational Inference
• Gradient Estimation Using Stochastic Computation Graphs
• Scaling Exact Inference for Discrete Probabilistic Programs
• Monte Carlo Gradient Estimation in Machine Learning
• Automatic Variational Inference in Stan
• Denotational validation of higher-order Bayesian inference
• Pathwise Derivatives Beyond the Reparameterization Trick
• Operator Variational Inference
• Variational inference with a quantum computer
• SPPL: Probabilistic Programming with Fast Exact Symbolic Inference
• Variational Bayes survival analysis for unemployment modelling
• The Generalized Reparameterization Gradient
• Black-Box Policy Search with Probabilistic Programs
• Design and Implementation of Probabilistic Programming Language Anglican
• Robust Variational Autoencoder
• When and How to Lift the Lockdown? Global COVID-19 Scenario Analysis and Policy Assessment using Compartmental Gaussian Processes
• Overdispersed Black-Box Variational Inference
• DiCE: The Infinitely Differentiable Monte-Carlo Estimator
• Automatic structured variational inference
• DisARM: An Antithetic Gradient Estimator for Binary Latent Variables
• Learning Proposals for Probabilistic Programs with Inference Combinators
• Effect Handling for Composable Program Transformations in Edward2
• Divide, Conquer, and Combine: a New Inference Strategy for Probabilistic Programs with Stochastic Support
• Addressing Variance Shrinkage in Variational Autoencoders using Quantile Regression
• Provable Smoothness Guarantees for Black-Box Variational Inference
• Amortized Rejection Sampling in Universal Probabilistic Programming
• Variational Rejection Sampling
• Automatic variational inference with cascading flows
• Applications of Probabilistic Programming (Master's thesis, 2015)
• Approximate Inference in Discrete Distributions with Monte Carlo Tree Search and Value Functions
• Spreadsheet Probabilistic Programming
• TyXe: Pyro-based Bayesian neural nets for Pytorch
• Pathwise Derivatives for Multivariate Distributions
• Encoding Domain Information with Sparse Priors for Inferring Explainable Latent Variables
• Stochastically Differentiable Probabilistic Programs
• Training Multimodal Systems for Classification with Multiple Objectives
• Towards Verified Stochastic Variational Inference for Probabilistic Programs