Inducing Probabilistic Programs by Bayesian Program Merging
By Irvin Hwang, Andreas Stuhlmüller, Noah D. Goodman. 2011
34 citations recorded by the source corpus.
This report outlines an approach to learning generative models from data. We express models as probabilistic programs, which allows us to capture abstract patterns within the examples. By choosing our language for programs to be an extension of the algebraic data type of the examples, we can begin with a program that generates all and only the examples. We then introduce greater abstraction, and hence generalization, incrementally to the extent that it improves the posterior probability of the examples given the program. Motivated by previous approaches to model merging and program induction, we search for such explanatory abstractions using program transformations. We consider two types of transformation: Abstraction merges common subexpressions within a program into new functions (a form of anti-unification). Deargumentation simplifies functions by reducing the number of arguments. We demonstrate that this approach finds key patterns in the domain of nested lists, including parameterized sub-functions and stochastic recursion.
Subfields: Reinforcement Learning, Probabilistic & Bayesian Methods
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Reinforcement Learning
In Probabilistic & Bayesian Methods
- Yura N. Perov
- Frank D. Wood
- Hongseok Yang
- Jan-Willem van de Meent
- Brooks Paige
- David Tolpin
- Zoubin Ghahramani
- Tuan Anh Le
Citations in the corpus
- An Introduction to Probabilistic Programming (2018, background): “This is the central idea in the work of Perov and Wood (2016) whose use of probabilistic programming is what distinguishes their work from the related literature (Gulwani et al., 2017; Hwang et al., 2011; Liang et al., 2010).”
- Automatic Sampler Discovery via Probabilistic Programming and Approximate Bayesian Computation (2016, background): “By this parameterising compilation links our approach to that of [13].”