[ICAPS] CFP: Third Workshop on Bridging the Gap Between AI Planning and Reinforcement Learning (PRL)

[Michael Katz]

================================================================
                       CALL FOR PAPERS

Bridging the Gap Between AI Planning and Reinforcement Learning
                           (PRL 2022)

               http://prl-theworkshop.github.io
                Singapore -  June 20-21, 2022.
================================================================


While  AI  Planning and  Reinforcement  Learning  communities focus  on  similar
sequential decision-making  problems, these communities remain  somewhat unaware
of each other  on specific problems, techniques,  methodologies, and evaluation.

This  workshop  aims  to  encourage discussion  and  collaboration  between  the
researchers in the  fields of AI planning and reinforcement  learning. We aim to
bridge  the  gap between  the  two  communities,  facilitate the  discussion  of
differences and similarities in existing techniques, and encourage collaboration
across the  fields. We  solicit interest  from AI researchers  that work  in the
intersection of planning  and reinforcement learning, in  particular, those that
focus on intelligent decision making. As  such, the joint workshop program is an
excellent  opportunity  to  gather  a  large and  diverse  group  of  interested
researchers.

Workshop topics:
================

The  workshop solicits work at the intersection of  the fields of  reinforcement
learning  and  planning.  One example  is so-called  goal-directed reinforcement
learning,  where a  goal must be  achieved,  and no partial credit  is given for
getting  closer to the goal.  In this case,  a usual metric is  success rate. We
also solicit work solely in one area that can influence advances in the other so
long as the connections are *clearly articulated* in the submission.

Submissions are invited for topics on, but not limited to:

 * Goal-directed reinforcement learning (model-based, Bayesian, deep, etc.)
 * Safe Reinforcement Learning and Planning
 * Monte Carlo Planning
 * Learning search heuristics for planner guidance
 * Model representation and learning for planning
 * Planning using approximated/uncertain (learned) models
 * Theoretical aspects of planning and reinforcement learning
 * Action policy analysis or certification
 * Reinforcement Learning and planning competition(s)
 * Applications of both reinforcement learning and planning
 * Various levels of generalization (across goals, objects/domain, domains)
 * Goal-oriented sequential decision methods combining planning, RL
   or other ML methods.


Important Dates:
================

 * Submission deadline: Friday 18 March, 2022 (UTC-12 timezone)
 * Notification date: Friday April 15th, 2022
 * Camera-ready deadline: Friday 10 June, 2022
 * Workshop date: June 20 or 21 (TBD), 2022

Submission Procedure:
=====================

We  solicit  workshop paper  submissions  relevant  to  the  above call  of  the
following types:

 * Long papers -- up to 8 pages + unlimited references / appendices
 * Short papers -- up to 4 pages + unlimited references / appendices
 * Extended abstracts -- up to 2 pages + unlimited references / appendices

Please format submissions  in AAAI style (see instructions in  the Author Kit at
https://www.aaai.org/Publications/Templates/AuthorKit22.zip). Authors considering
submitting to the workshop papers rejected from other conferences, please ensure
you  do  your  utmost to  address  the  comments given  by the reviewers. Please
do  not submit papers that are already accepted for the main ICAPS conference to
the workshop.

Some accepted long  papers will be accepted as contributed  talks.  All accepted
long and short papers and extended abstracts  will be given a slot in the poster
presentation session.   Extended abstracts  are intended  as brief  summaries of
already published papers,  preliminary work, position papers  or challenges that
might help bridge the gap.

As the main purpose  of this workshop is to solicit  discussion, the authors are
invited to use the appendix of their submissions for that purpose.


Paper submissions should be made through EasyChair:
  https://easychair.org/my/conference?conf=prl20220

Workshop Organizers
=====================
Michael Katz, IBM T.J. Watson Research Center, NY, USA
Hector Palacios, ServiceNow Research, Montreal, Canada
Vicenç Gómez, Universitat Pompeu Fabra, Barcelona, Spain


Please send your inquiries to prl.the...@gmail.com