Guardado en:
Detalles Bibliográficos
Autores principales: Tuero, Jake, Buro, Michael, Lelis, Levi H. S.
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2506.07255
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915647533875200
author Tuero, Jake
Buro, Michael
Lelis, Levi H. S.
author_facet Tuero, Jake
Buro, Michael
Lelis, Levi H. S.
contents Policy tree search is a family of tree search algorithms that use a policy to guide the search. These algorithms provide guarantees on the number of expansions required to solve a given problem that are based on the quality of the policy. While these algorithms have shown promising results, the process in which they are trained requires complete solution trajectories to train the policy. Search trajectories are obtained during a trial-and-error search process. When the training problem instances are hard, learning can be prohibitively costly, especially when starting from a randomly initialized policy. As a result, search samples are wasted in failed attempts to solve these hard instances. This paper introduces a novel method for learning subgoal-based policies for policy tree search algorithms. The subgoals and policies conditioned on subgoals are learned from the trees that the search expands while attempting to solve problems, including the search trees of failed attempts. We empirically show that our policy formulation and training method improve the sample efficiency of learning a policy and heuristic function in this online setting.
format Preprint
id arxiv_https___arxiv_org_abs_2506_07255
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Subgoal-Guided Policy Heuristic Search with Learned Subgoals
Tuero, Jake
Buro, Michael
Lelis, Levi H. S.
Artificial Intelligence
Policy tree search is a family of tree search algorithms that use a policy to guide the search. These algorithms provide guarantees on the number of expansions required to solve a given problem that are based on the quality of the policy. While these algorithms have shown promising results, the process in which they are trained requires complete solution trajectories to train the policy. Search trajectories are obtained during a trial-and-error search process. When the training problem instances are hard, learning can be prohibitively costly, especially when starting from a randomly initialized policy. As a result, search samples are wasted in failed attempts to solve these hard instances. This paper introduces a novel method for learning subgoal-based policies for policy tree search algorithms. The subgoals and policies conditioned on subgoals are learned from the trees that the search expands while attempting to solve problems, including the search trees of failed attempts. We empirically show that our policy formulation and training method improve the sample efficiency of learning a policy and heuristic function in this online setting.
title Subgoal-Guided Policy Heuristic Search with Learned Subgoals
topic Artificial Intelligence
url https://arxiv.org/abs/2506.07255