no code implementations • 14 Jul 2019 • Feras A. Saad, Marco F. Cusumano-Towner, Ulrich Schaechtle, Martin C. Rinard, Vikash K. Mansinghka
These techniques work with probabilistic domain-specific data modeling languages that capture key properties of a broad class of data generating processes, using Bayesian inference to synthesize probabilistic programs in these modeling languages given observed data.
no code implementations • 11 Jan 2018 • Marco F. Cusumano-Towner, Vikash K. Mansinghka
Monte Carlo inference has asymptotic guarantees, but can be slow when using generic proposals.
no code implementations • NeurIPS 2017 • Marco F. Cusumano-Towner, Vikash K. Mansinghka
This paper introduces the auxiliary inference divergence estimator (AIDE), an algorithm for measuring the accuracy of approximate inference algorithms.
1 code implementation • 17 Apr 2017 • Marco F. Cusumano-Towner, Alexey Radul, David Wingate, Vikash K. Mansinghka
Intelligent systems sometimes need to infer the probable goals of people, cars, and robots, based on partial observations of their motion.
no code implementations • 14 Dec 2016 • Marco F. Cusumano-Towner, Vikash K. Mansinghka
This paper introduces the probabilistic module interface, which allows encapsulation of complex probabilistic models with latent variables alongside custom stochastic approximate inference machinery, and provides a platform-agnostic abstraction barrier separating the model internals from the host probabilistic inference system.
no code implementations • 7 Dec 2016 • Marco F. Cusumano-Towner, Vikash K. Mansinghka
A key limitation of sampling algorithms for approximate inference is that it is difficult to quantify their approximation error.
no code implementations • 31 May 2016 • Marco F. Cusumano-Towner, Vikash K. Mansinghka
This paper introduces a new technique for quantifying the approximation error of a broad class of probabilistic inference programs, including ones based on both variational and Monte Carlo approaches.