Bayesian Synthesis of Probabilistic Programs for Automatic Data Modeling

This paper presents new techniques for automatically constructing probabilistic programs for data analysis, interpretation, and prediction. 

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.

It uses generative programming techniques to produce domain specific code - see below:

https://dl.acm.org/ft_gateway.cfm?id=3290350&ftid=2027616&dwn=1&CFID=46105614&CFTOKEN=bd554558d8f7ab57-41952796-0E75-1855-A8EFED89006D428F