Planning and reinforcement learning are two key approaches to sequential decision making. Multi-step approximate real-time dynamic programming, a recently successful algorithm class of which AlphaZero [Silver et al., 2018] is an example, combines both by nesting planning within a learning loop. However, the combination of planning and learning introduces a new question: how should we balance time spend on planning, learning and acting? The importance of this trade-off has not been explicitly studied before. We show that it is actually of key importance, with computational results indicating that we should neither plan too long nor too short. Conceptually, we identify a new spectrum of planning-learning algorithms which ranges from exhaustive search (long planning) to model-free RL (no planning), with optimal performance achieved midway.
|Title of host publication||ICAPS: PRL 2020|
|Subtitle of host publication||Proceedings of the 1st Workshop on Bridging the Gap Between AI Planning and Reinforcement Learning (PRL)|
|Editors||Alan Fern, Vicenc Gomez, Anders Jonsson, Michael Katz, Hector Palacios, Scott Sanner|
|Publisher||Association for the Advancement of Artificial Intelligence (AAAI)|
|Number of pages||8|
|Publication status||Published - 2020|
|Event||ICAPS 2020: 30th International Conference on Automated Planning and Scheduling - Virtual Nancy, France|
Duration: 19 Oct 2020 → 30 Oct 2020
Conference number: 30th
|Period||19/10/20 → 30/10/20|
Bibliographical noteGreen Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care
Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.