Privileged information dropout in reinforcement learning
File(s)
Accepted version
Author(s)
Kamienny, Pierre-Alexandre
Arulkumaran, Kai
Behbahani, Feryal
Boehmer, Wendelin
Whiteson, Shimon
Type
Working Paper
Abstract
Using privileged information during training can improve the sample
efficiency and performance of machine learning systems. This paradigm has been
applied to reinforcement learning (RL), primarily in the form of distillation
or auxiliary tasks, and less commonly in the form of augmenting the inputs of
agents. In this work, we investigate Privileged Information Dropout (\pid) for
achieving the latter which can be applied equally to value-based and
policy-based RL algorithms. Within a simple partially-observed environment, we
demonstrate that \pid outperforms alternatives for leveraging privileged
information, including distillation and auxiliary tasks, and can successfully
utilise different types of privileged information. Finally, we analyse its
effect on the learned representations.
efficiency and performance of machine learning systems. This paradigm has been
applied to reinforcement learning (RL), primarily in the form of distillation
or auxiliary tasks, and less commonly in the form of augmenting the inputs of
agents. In this work, we investigate Privileged Information Dropout (\pid) for
achieving the latter which can be applied equally to value-based and
policy-based RL algorithms. Within a simple partially-observed environment, we
demonstrate that \pid outperforms alternatives for leveraging privileged
information, including distillation and auxiliary tasks, and can successfully
utilise different types of privileged information. Finally, we analyse its
effect on the learned representations.
Date Issued
2020-05-19
Citation
2020
Publisher
arXiv
Copyright Statement
© 2020 The Author(s)
Identifier
http://arxiv.org/abs/2005.09220v1
Subjects
cs.LG
cs.LG
cs.AI
stat.ML
Publication Status
Published