From conversational agents to online recommendation to search and advertising, we are already interacting with increasingly sophisticated sequential decision making systems in daily life. Traditionally, sequential decision making research has focused on balancing the exploration-exploitation trade-off, or casting the interaction paradigm under reinforcement / imitation learning dichotomy. We aim to take a holistic view and call for a collective effort to translate principled research ideas into practically relevant solutions in both existing and new domains, such as healthcare, education, safe autonomous vehicles and robots, etc.
This workshop aims to bring together researchers from industry and academia in order to describe recent advances and discuss future research directions pertaining to real-world sequential decision making, broadly construed. We aim to highlight new and emerging research opportunities for the machine learning community that arise from the evolving needs for making decision making theoretically and practically relevant for realistic applications.
Research interest in reinforcement and imitation learning has surged significantly over the past several years, with the empirical successes of self-playing in games and availability of increasingly realistic simulation environments. We believe the time is ripe for the research community to push beyond simulated domains and start exploring research directions that directly address the real-world need for optimal decision making. We are particularly interested in understanding the current theoretical and practical challenges that prevent broader adoption of current policy learning and evaluation algorithms in high-impact applications, across a broad range of domains.
|8:45 - 9:00||Opening Remarks|
|9:00 - 9:30||Invited Talk|
|9:30 - 10:00||Invited Talk|
|10:00 - 10:30||Morning Coffee Break and Poster Session|
|10:30 - 11:00||Invited Talk|
|11:00 - 11:30||Invited Talk|
|11:30 - 12:00||Poster Session|
|12:00 - 12:30||Invited Talk|
|12:30 - 14:00||Lunch and Closing Remarks|
The list of accepted papers will be posted after the author notification date.
Papers submitted to the workshop should be up to five pages long excluding references and appendix, and in ICML 2019 format. As the review process is not blind, authors can reveal their identity in their submissions. All inquiries may be sent to RealWorldSDM@gmail.com .
Submissions page: Real-world Sequential Decision Making Workshop 2019.
We invite researchers to submit both theoretical and applied work along several possible dimensions:
Application papers: applications of learning-based techniques to real-life sequential decision making (robotics, healthcare, education, transportation & energy, smart-grids, sustainability, NLP, social media & advertising, agriculture, manufacturing, economics and policy)
Method papers that address real-world desiderata and concerns: safety, reliable decision making, theoretical guarantees, verifiability & interpretability, data-efficiency, data-heterogeneity, efficient exploration, counterfactual reasoning and off-policy evaluation, cost function design, efficient implementations to large-scale systems
Cross-boundary papers along the theme of RL+X where X indicate areas not commonly viewed as RL in contemporary research. We would like to encourage researchers to explore the interface between traditional RL with:
Other related areas in machine learning including but not limited to: imitation learning, transfer learning, active learning, structured prediction, off-policy learning, fairness in ML, privacy
Areas outside of machine learning including but not limited to: control theory & dynamical systems, formal methods, causal inference, game-theory, operations research, systems research, human-computer interactions, human behavior modeling
Paper Submission Deadline: May 20, 2019, 11:59 PM PST
Author Notification: May 26, 2019, 11:59 PM PST
Final Version: June 13, 2019, 11:59 PM PST
Workshop: June 14 or 15, 2019