Safe Learning of Lifted Action Models

Brendan Juba, Hai S. Le, Roni Stern

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Creating a domain model, even for classical, domain-independent planning, is a notoriously hard knowledge-engineering task. A natural approach to solve this problem is to learn a domain model from observations. However, model learning approaches frequently do not provide safety guarantees: the learned model may assume actions are applicable when they are not, and may incorrectly capture actions' effects. This may result in generating plans that will fail when executed. In some domains such failures are not acceptable, due to the cost of failure or inability to replan online after failure. In such settings, all learning must be done offline, based on some observations collected, e.g., by some other agents or a human. Through this learning, the task is to generate a plan that is guaranteed to be successful. This is called the model-free planning problem. Prior work proposed an algorithm for solving the model-free planning problem in classical planning. However, they were limited to learning grounded domains, and thus they could not scale. We generalize this prior work and propose the first safe model-free planning algorithm for lifted domains. We prove the correctness of our approach, and provide a statistical analysis showing that the number of trajectories needed to solve future problems with high probability is linear in the potential size of the domain model. We also present experiments on twelve IPC domains showing that our approach is able to learn the real action model in all cases with at most two trajectories.

Original languageAmerican English
Title of host publicationProceedings of the 18th International Conference on Principles of Knowledge Representation and Reasoning, KR 2021
EditorsMeghyn Bienvenu, Gerhard Lakemeyer, Esra Erdem
Pages379-389
Number of pages11
ISBN (Electronic)9781956792997
StatePublished - 1 Jan 2021
Event18th International Conference on Principles of Knowledge Representation and Reasoning, KR 2021 - Virtual, Online
Duration: 3 Nov 202112 Nov 2021

Publication series

NameProceedings of the 18th International Conference on Principles of Knowledge Representation and Reasoning, KR 2021

Conference

Conference18th International Conference on Principles of Knowledge Representation and Reasoning, KR 2021
CityVirtual, Online
Period3/11/2112/11/21

All Science Journal Classification (ASJC) codes

  • Software
  • Logic

Fingerprint

Dive into the research topics of 'Safe Learning of Lifted Action Models'. Together they form a unique fingerprint.

Cite this