PROBPROG 2018

The International Conference on Probabilistic Programming

probprog-banner

Program Overview

Thursday October 4th

The Industry Day will take place in the McGovern Institute for Brain and Cognitive Science at MIT.

Start End Activity Location

10:00

Registration

BCS 3310

10:30

10:40

Welcome

BCS 3310

10:40

12:00

Speed Networking

BCS 3310

12:00

13:00

Lunch Break

13:00

14:30

Tutorial by Vikash Mansinghka

BCS 3189

14:30

15:00

Coffee Break and Networking

BCS 3310

15:00

16:30

Industry Discussions
Probabilistic Programming for Analytics Consulting
Probabilistic Programming and Deep Learning
(Email us to propose additional topics)

BCS 3189

15:00

16:30

Developer Meetup
Anglican: David Tolpin, Jan-Willem van de Meent, Frank Wood
BayesDB: Ulrich Schaechtle, Feras Saad
Birch: Lawrence Murray
BLOG: Yi Wu
ForneyLab.jl: Bert de Vries, Marco Cox, Thijs van de Laar
Gen: Marco Cusumano-Towner
Infer.NET: Yordan Zaykov
MetaProb: Zane Shelby, Tim Trautman, Jonathan Rees, Alex Lew
MonadBayes: Adam Ścibior
Omega: Javier Burroni
Probabilistic Torch: Alican Bozkurt, Jan-Willem van de Meent
PSL: Eriq Augustine
PyMC3: Colin Carroll
Pyro: Eli Bingham, Fritz Obermeyer
Rainier: Avi Bryant
Stan: Matthijs Vakar, Sean Talts
TensorFlow Probability: Alexey Radul, Dave Moore Turing.jl: Hong Ge

BCS 3310

Friday October 5th

The Practice of Probabilistic Programming and Statistics and Data Analysis session will take place in MIT Wiesner Building E15.

Start End Activity Location

8:00

Registration

Lower Atrium

8:40

12:00

Session: Practice of Probabilistic Programming

Bartos Theater

8:40

9:20

Keynote: Zoubin Ghahramani (Uber AI Labs, University of Cambridge)

Probabilistic Machine Learning: From theory to industrial impact

9:20

9:40

Talk: Daniel Lee (Stan Development Team, Generable)

Dear Stan, I meant to write you sooner but I just been busy

9:40

10:00

Talk: Yordan Zaykov (Microsoft Research)

Probabilistic programming in production with Infer.NET

10:00

10:20

Talk: Michael Tingley (Facebook)

Probabilistic programming @ FB

10:20

10:40

Coffee Break

10:40

11:00

Talk: Daniel Ritchie (Brown)

Probabilistic Programming for Computer Graphics

11:00

11:20

Talk: Dustin Tran (Google)

What might deep learners learn from probabilistic programming?

11:20

11:40

Talk: Brooks Paige (Alan Turing Institute, University of Cambridge)

Semi-interpretable probabilistic models

11:40

12:00

Talk: Ulrich Schaechtle (MIT)

Automated data modeling for science via Bayesian probabilistic program synthesis

12:00

15:00

Lunch and Poster Session

15:00

18:00

Session: Statistics and Data Analysis

Bartos Theater

15:00

15:40

Keynote: Dave Blei (Columbia University)

Black Box Variational Inference

15:40

16:00

Talk: Lawrence Murray (Uppsala University)

Automated learning with a probabilistic programming language: Birch

16:00

16:20

Talk: Dan Roy (University of Toronto)

Algorithmic Barriers to Representing Conditional Independence

16:20

16:40

Coffee Break

16:40

17:00

Talk: Tom Rainforth (University of Oxford)

Nesting Probabilistic Programs

17:00

17:20

Talk: Maria Gorinova (University of Edinburgh)

SlicStan: Optimising Probabilistic Programs using Information Flow Analysis

17:20

17:40

Talk: Hong Ge & Kai Xu (University of Cambridge)

The Turing Language for Probabilistic Programming

17:40

18:00

Talk: Feras Saad (MIT)

Bayesian Synthesis of Probabilistic Programs for Automatic Data Modeling

Saturday October 6th

The Probabilistic Programming and Intelligence and Languages and Systems sessions will take place in MIT Wiesner Building E15.

Start End Activity Location

8:00

Registration

Lower Atrium

8:40

12:00

Session: Intelligence

Bartos Theater

8:40

9:20

Keynote: Stuart Russell (UC Berkeley)

Probabilistic programming and AI

9:20

9:40

Talk: Josh Tenenbaum (MIT)

9:40

10:00

Talk: Kristian Kersting (TU Darmstadt)

Democratizing Machine Learning using Probabilistic Programming

10:00

10:20

Talk: Frank Wood (University of British Columbia)

Inference Compilation

10:20

10:40

Coffee Break

10:40

11:00

Community Announcements and Discussion

11:00

11:20

Talk: Noah Goodman (Uber AI Labs, Stanford)

Accelerating science with PPLs for experiment design

11:20

11:40

Talk: Marco Cusumano-Towner (MIT)

Gen: A Flexible System for Programming Probabilistic AI

11:40

12:00

Talk: Kevin Ellis (MIT)

Growing Libraries of Subroutines with Wake/Sleep Bayesian Program Learning

12:00

14:40

Lunch and Poster Session

14:40

17:40

Session: Languages and Systems

Bartos Theater

14:40

15:20

Keynote: Jean-Baptiste Tristan (Oracle Labs)

Compilation of Probabilistic Programs

15:20

15:40

Talk: Angelika Kimmig (University of Cardiff)

A short introduction to probabilistic logic programming

15:40

16:00

Talk: Hongseok Yang (KAIST)

Reparameterization Gradient for Non-differentiable Models from Probabilistic Programming

16:00

16:20

Talk: Chung-Chieh (Ken) Shan (Indiana University)

Calculating distributions

16:20

16:40

Coffee Break

16:40

17:00

Talk: Adam Scibior (University of Cambridge)

Denotational account of approximate Bayesian inference

17:00

17:20

Talk: Sarah Chasins (UC Berkeley)

Data-Driven Synthesis of Full Probabilistic Programs

17:20

17:40

Talk: Eric Atkinson (MIT)

Verifying Handcoded Probabilistic Inference Procedures

17:40

18:00

Talk: Timon Gehr (ETH)

PSI: Exact Symbolic Inference for Probabilistic Programs

Sponsors

Platinum Sponsor:

google-logo

Gold Sponsor:

intel-logo