Learning to Achieve Goals
Proceedings of the Thirteenth International Joint Conference on Artificial Intelligence. Machine Learning. Pages 1094-1098.
Abstract
Temporal difference methods solve the temporal credit assignment problem for reinforcement learning. An important subproblem of general reinforcement learning is learning to achieve dynamic goals. Although existing temporal difference methods, such as Q learning, can be applied to this problem , they do not take advantage of its special structure. This paper presents the DG-learning algorithm, which learns efficiently to achie ve dynamically changing goals and exhibits good knowledge transfer between goals. In addition, this paper shows how traditional relaxation techniques can be applied to the problem . Finally, experimental results ar e given that demonstrate the superiority of DG learning over Q learning in a moderately large, synthetic , non-deterministic domain .
@inproceedings{ijcai1993p151,
title = {Learning to Achieve Goals},
author = {Kaelbling, Leslie Pack},
booktitle = {Proceedings of the Thirteenth International Joint Conference on
Artificial Intelligence, {IJCAI-93}},
publisher = {International Joint Conferences on Artificial Intelligence},
editor = {Ruzena Bajcsy},
pages = {1094--1098},
year = {1993},
note = {Machine Learning},
url = {https://ijcai.org/proceedings/1993/151},
}
TY - CPAPER TI - Learning to Achieve Goals AU - Kaelbling, Leslie Pack T2 - Proceedings of the Thirteenth International Joint Conference on Artificial Intelligence, IJCAI-93 PB - International Joint Conferences on Artificial Intelligence PY - 1993 SP - 1094 EP - 1098 UR - https://ijcai.org/proceedings/1993/151 ER -Copied
Machine-readable versions of this paper, for tools and AI agents:
