Tue 16 Jun 2015 09:40 - 10:05 at PLDI Main RED (Portland 256) - Synthesis I Chair(s): Kathleen Fisher

We show how to automatically synthesize probabilistic programs from real-world datasets. Such a synthesis is feasible due to a combination of two techniques: (1) We borrow the idea of “sketching” from synthesis of deterministic programs, and allow the programmer to write a skeleton program with “holes”. Sketches enable the programmer to communicate domain-specific intuition about the structure of the desired program and prune the search space, and (2) we design an efficient Markov Chain Monte Carlo (MCMC) based synthesis algorithm to instantiate the holes in the sketch with program fragments. Our algorithm efficiently synthesizes a probabilistic program that is most consistent with the data. A core difficulty in synthesizing probabilistic programs is computing the likelihood L(P | D) of a candidate program P generating data D. We propose an approximate method to compute likelihoods using mixtures of Gaussian distributions, thereby avoiding expensive computation of integrals. The use of such approximations enables us to speed up evaluation of the likelihood of candidate programs by a factor of 1000, and makes Markov Chain Monte Carlo based search feasible. We have implemented our algorithm in a tool called PSKETCH, and our results are encouraging—PSKETCH is able to automatically synthesize 16 non-trivial real-world probabilistic programs.

Tue 16 Jun

Displayed time zone: Tijuana, Baja California change

09:15 - 10:55
Synthesis IResearch Papers at PLDI Main RED (Portland 256)
Chair(s): Kathleen Fisher Tufts University
09:15
25m
Talk
Efficient Synthesis of Network Updates
Research Papers
Jedidiah McClurg University of Colorado Boulder, Hossein Hojjat Cornell University, Pavol Cerny University of Colorado Boulder, Nate Foster Cornell University
Pre-print Media Attached
09:40
25m
Talk
Efficient Synthesis of Probabilistic Programs
Research Papers
Aditya Nori Microsoft Research, UK, Sherjil Ozair IIT Delhi, Sriram Rajamani Microsoft Research, Deepak Vijaykeerthy Microsoft Research
Media Attached
10:05
25m
Talk
FlashRelate: Extracting Relational Data from Semi-Structured Spreadsheets Using Examples
Research Papers
Dan Barowy University of Massachusetts Amherst, Sumit Gulwani Microsoft Research, Ted Hart Microsoft Research, Benjamin Zorn Microsoft Research
Media Attached
10:30
25m
Talk
Synthesizing Data Structure Transformations from Input-Output Examples
Research Papers
Jack Feser Rice University, Swarat Chaudhuri Rice University, Işıl Dillig University of Texas, Austin
Media Attached