Dear members of AI Inference and Machine Learning Communities,
After several years of slumber, we are pleased to announce the resurrection of the UAI inference competition. This year we introduce a new machine learning component to the competition (read more below).
The goal of this competition is to explore our communities’ various approaches to approximating quantities of graphical models such as summation queries of the partition function or probability of evidence and optimization queries of MAP or marginal MAP, evaluating the algorithms on an array of problems with varying difficulties. As alluded to, this year we also include a new category that hosts the machine learning challenge: multi-label classification (MLC), extending the competition to exploring various learning algorithms. We hope this event will continue to stimulate algorithms development and especially assess the state of the art in solvers for inference and learning in graphical models.
The timeline for the competition is as follows:
June, 2022
sample benchmarks to be released
solver submission opens (June 15th)
July, 2022
solver evaluations begin
August, 2022
mid-competition update (to be announced at UAI-2022)
October, 2022
solver submissions close
final competition results announced!
More information can be found on this year’s competition website: https://uaicompetition.github.io/uci-2022/
We encourage anyone interested - whether new to the field or veterans - to join! Please reach out to us via uaicomp...@gmail.com with any questions.
We look forward to hearing from you and including your solvers in the competition!
UAI-2022 Competition Organizers
Rina Dechter (University of California, Irvine)
Vibhav Gogate (University of Texas, Dallas)
Alexander Ihler (University of California, Irvine)
Junkyu Lee (IBM Research)
Bobak Pezeshki (University of California, Irvine)
-- Rina Dechter Distinguished Professor, Department of Computer Science University of California, Irvine Irvine, CA 92697-3425 (949) 824-6556 dec...@ics.uci.edu http://www.ics.uci.edu/~dechter