Past Meetups (26)

What we're about

Take a look at material from past meetups in our associated Github Repo (https://github.com/suhailshergill/ProPL-meetup).

Also come discuss ideas with like-minded individuals in our Gitter Chatroom (https://gitter.im/suhailshergill/ProPL-meetup).

This group will focus on recent advances in Probabilistic Programming. Paraphrasing from the wikipedia page (https://en.wikipedia.org/wiki/Probabilistic_programming_language) Probabilistic Programming represents an attempts to unify general purpose programming with the process of probabilistic modelling. Discussions in this meetup will be focused on understanding research papers, to studying opensource implementations, to sharing our experiences with using existing probabilistic programming frameworks such as Stan, PyMC, Church etc. with other enthusiasts.

To those wondering why we should care about Probabilistic Programming, the one word answer is: expressivity. To elaborate, as demonstrated in a 2015 paper in Computer Vision and Pattern Recognition the authors were able to express in 50 lines of code of a probabilistic programming language, what otherwise would ordinarily take thousands of lines of code in a conventional programming language (https://en.wikipedia.org/wiki/Probabilistic... (https://en.wikipedia.org/wiki/Probabilistic_programming_language#Applications)).

Some useful links:

• Why Probabilistic Programming Matters (http://zinkov.com/posts/2012-06-27-why-prob-programming-matters/)

• Wikipedia (https://en.wikipedia.org/wiki/Probabilistic_programming_language): General description of Probabilistic Programming and links to various implementations

• Probabilistic Models of Cognition (https://probmods.org/): Good Tutorial to understand the essence of Probabilistic Programming and Modelling using examples implemented in Church.

• The Design and Implementation of Probabilistic Programming Languages (http://dippl.org/): This online book explains how to implement PPLs by lightweight embedding into a host language. It illustrates this by designing and implementing WebPPL, a small PPL embedded in Javascript. The book shows how to implement several algorithms for universal probabilistic inference, including priority-based enumeration with caching, particle filtering, and Markov chain Monte Carlo.

• Church Probabilistic Programming Language wiki (http://projects.csail.mit.edu/church/wiki/Church)

NOTE: We are always looking for sponsors who would be willing to lend a hand in hosting meetups. If you or your organization would like to help out, please contact the organizer.

Members (477)

Photos (46)