TY - GEN
T1 - Learning Safe Numeric Action Models
AU - Mordoch, Argaman
AU - Juba, Brendan
AU - Stern, Roni
N1 - Publisher Copyright: Copyright © 2023, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
PY - 2023/6/27
Y1 - 2023/6/27
N2 - Powerful domain-independent planners have been developed to solve various types of planning problems. These planners often require a model of the acting agent’s actions, given in some planning domain description language. Yet obtaining such an action model is a notoriously hard task. This task is even more challenging in mission-critical domains, where a trial-and-error approach to learning how to act is not an option. In such domains, the action model used to generate plans must be safe, in the sense that plans generated with it must be applicable and achieve their goals. Learning safe action models for planning has been recently explored for domains in which states are sufficiently described with Boolean variables. In this work, we go beyond this limitation and propose the Numeric Safe Action Model Learning (N-SAM) algorithm. N-SAM runs in time that is polynomial in the number of observations and, under certain conditions, is guaranteed to return safe action models. We analyze its worst-case sample complexity, which may be intractable for some domains. Empirically, however, N-SAM can quickly learn a safe action model that can solve most problems in the domain.
AB - Powerful domain-independent planners have been developed to solve various types of planning problems. These planners often require a model of the acting agent’s actions, given in some planning domain description language. Yet obtaining such an action model is a notoriously hard task. This task is even more challenging in mission-critical domains, where a trial-and-error approach to learning how to act is not an option. In such domains, the action model used to generate plans must be safe, in the sense that plans generated with it must be applicable and achieve their goals. Learning safe action models for planning has been recently explored for domains in which states are sufficiently described with Boolean variables. In this work, we go beyond this limitation and propose the Numeric Safe Action Model Learning (N-SAM) algorithm. N-SAM runs in time that is polynomial in the number of observations and, under certain conditions, is guaranteed to return safe action models. We analyze its worst-case sample complexity, which may be intractable for some domains. Empirically, however, N-SAM can quickly learn a safe action model that can solve most problems in the domain.
UR - http://www.scopus.com/inward/record.url?scp=85168253204&partnerID=8YFLogxK
M3 - Conference contribution
T3 - Proceedings of the 37th AAAI Conference on Artificial Intelligence, AAAI 2023
SP - 12079
EP - 12086
BT - AAAI-23 Technical Tracks 10
A2 - Williams, Brian
A2 - Chen, Yiling
A2 - Neville, Jennifer
T2 - 37th AAAI Conference on Artificial Intelligence, AAAI 2023
Y2 - 7 February 2023 through 14 February 2023
ER -