https://sites.google.com/view/tpm2021/home

There is an increasing need for probabilistic machine learning (ML) models that are able to deliver probabilistic inference with guarantees (reliability) while allowing to flexibly represent complex real-world scenarios (expressiveness). This edition of the workshop on tractable probabilistic models (TPMs) aims at bringing together researches working on different fronts of this trade-off between reliable and expressive models in modern probabilistic ML.

Recent years have shown how TPMs can achieve such a sensible trade-off in tasks like image classification, completion and generation, activity recognition, language and speech modeling, bioinformatics, verification and diagnosis of physical systems, to name but a few. Examples of TPMs comprise – but are not limited to – i) neural autoregressive models; ii) normalizing flows; iii) bounded-treewidth probabilistic graphical models (PGMs); iv) determinantal point processes; v) PGMs with high girth or weak potentials; vi) exchangeable probabilistic models and models exploiting symmetries and invariances and vii) probabilistic circuits (arithmetic circuits, sum-product networks, probabilistic sentential decision diagrams, cutset networks, etc.).

Topics

We especially encourage submissions highlighting the challenges and opportunities for tractable inference, including, but not limited to:

• New tractable representations in discrete, continuous and hybrid domains,
• Learning algorithms for TPMs
• Theoretical and empirical analysis of tractable models
• Connections between TPM classes
• TPMs for responsible, robust and explainable AI
• Retrospective works, tutorials, and surveys
• Approximate inference algorithms with guarantees
• Tractable neuro-symbolic and/or relational modeling
• Applications of tractable probabilistic modeling

 Submission Instructions

We invite three types of submissions:

• Original research papers: advances in TPM, not previously published in an archival conference or journal.
• Recently published research papers: advances in TPM, already published at a recent venue.
• Position papers (abstracts): discussing tendencies, issues or future venues of interest for the TPM community.

All submissions must be electronic (through the link below), and must closely follow the formatting guidelines at https://sites.google.com/view/tpm2021/call-for-papers. Reviewing for TPM 2021 is single-blind. We recommend that you refer to your prior work in the third person wherever possible. We also encourage links to public repositories such as github to share code and/or data.

Submission Link: https://openreview.net/group?id=auai.org/UAI/2021/Workshop/TPM

 ***Accepted papers will be considered for the best paper award***

 Important Dates

§ Paper submission deadline: May 28, 2021 AOE (UTC-12:00h)

§ Notification to authors: June 28, 2021

§ Camera-ready version: July 27, 2021 AOE (UTC-12:00h) *

§ Workshop Date: July 30, 2021

Organizers

Antonio Vergari (University of California, Los Angeles)

Tahrima Rahman (University of Texas, Dallas)

Robert Peharz (TU Eindhoven)

Alejandro Molina (TU Darmstadt)

Pedram Rooshenas (University of North Carolina, Charlotte)

Daniel Lowd  (University of Oregon)

Zoubin Ghahramani (Google AI) 

For any questions, contact us at tpmworkshop2021@gmail.com

***Please consider sharing this CFP in your network***