Classifying Options for Deep Reinforcement Learning
Machine Learning
2017-06-20 v3 Artificial Intelligence
Machine Learning
Abstract
In this paper we combine one method for hierarchical reinforcement learning - the options framework - with deep Q-networks (DQNs) through the use of different "option heads" on the policy network, and a supervisory network for choosing between the different options. We utilise our setup to investigate the effects of architectural constraints in subtasks with positive and negative transfer, across a range of network capacities. We empirically show that our augmented DQN has lower sample complexity when simultaneously learning subtasks with negative transfer, without degrading performance when learning subtasks with positive transfer.
Cite
@article{arxiv.1604.08153,
title = {Classifying Options for Deep Reinforcement Learning},
author = {Kai Arulkumaran and Nat Dilokthanakul and Murray Shanahan and Anil Anthony Bharath},
journal= {arXiv preprint arXiv:1604.08153},
year = {2017}
}
Comments
IJCAI 2016 Workshop on Deep Reinforcement Learning: Frontiers and Challenges