Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning

Junhyuk Oh, Satinder Singh, Honglak Lee, Pushmeet Kohli
Proceedings of the 34th International Conference on Machine Learning, PMLR 70:2661-2670, 2017.

Abstract

As a step towards developing zero-shot task generalization capabilities in reinforcement learning (RL), we introduce a new RL problem where the agent should learn to execute sequences of instructions after learning useful skills that solve subtasks. In this problem, we consider two types of generalizations: to previously unseen instructions and to longer sequences of instructions. For generalization over unseen instructions, we propose a new objective which encourages learning correspondences between similar subtasks by making analogies. For generalization over sequential instructions, we present a hierarchical architecture where a meta controller learns to use the acquired skills for executing the instructions. To deal with delayed reward, we propose a new neural architecture in the meta controller that learns when to update the subtask, which makes learning more efficient. Experimental results on a stochastic 3D domain show that the proposed ideas are crucial for generalization to longer instructions as well as unseen instructions.

Cite this Paper


BibTeX
@InProceedings{pmlr-v70-oh17a, title = {Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning}, author = {Junhyuk Oh and Satinder Singh and Honglak Lee and Pushmeet Kohli}, booktitle = {Proceedings of the 34th International Conference on Machine Learning}, pages = {2661--2670}, year = {2017}, editor = {Precup, Doina and Teh, Yee Whye}, volume = {70}, series = {Proceedings of Machine Learning Research}, month = {06--11 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v70/oh17a/oh17a.pdf}, url = { http://proceedings.mlr.press/v70/oh17a.html }, abstract = {As a step towards developing zero-shot task generalization capabilities in reinforcement learning (RL), we introduce a new RL problem where the agent should learn to execute sequences of instructions after learning useful skills that solve subtasks. In this problem, we consider two types of generalizations: to previously unseen instructions and to longer sequences of instructions. For generalization over unseen instructions, we propose a new objective which encourages learning correspondences between similar subtasks by making analogies. For generalization over sequential instructions, we present a hierarchical architecture where a meta controller learns to use the acquired skills for executing the instructions. To deal with delayed reward, we propose a new neural architecture in the meta controller that learns when to update the subtask, which makes learning more efficient. Experimental results on a stochastic 3D domain show that the proposed ideas are crucial for generalization to longer instructions as well as unseen instructions.} }
Endnote
%0 Conference Paper %T Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning %A Junhyuk Oh %A Satinder Singh %A Honglak Lee %A Pushmeet Kohli %B Proceedings of the 34th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2017 %E Doina Precup %E Yee Whye Teh %F pmlr-v70-oh17a %I PMLR %P 2661--2670 %U http://proceedings.mlr.press/v70/oh17a.html %V 70 %X As a step towards developing zero-shot task generalization capabilities in reinforcement learning (RL), we introduce a new RL problem where the agent should learn to execute sequences of instructions after learning useful skills that solve subtasks. In this problem, we consider two types of generalizations: to previously unseen instructions and to longer sequences of instructions. For generalization over unseen instructions, we propose a new objective which encourages learning correspondences between similar subtasks by making analogies. For generalization over sequential instructions, we present a hierarchical architecture where a meta controller learns to use the acquired skills for executing the instructions. To deal with delayed reward, we propose a new neural architecture in the meta controller that learns when to update the subtask, which makes learning more efficient. Experimental results on a stochastic 3D domain show that the proposed ideas are crucial for generalization to longer instructions as well as unseen instructions.
APA
Oh, J., Singh, S., Lee, H. & Kohli, P.. (2017). Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning. Proceedings of the 34th International Conference on Machine Learning, in Proceedings of Machine Learning Research 70:2661-2670 Available from http://proceedings.mlr.press/v70/oh17a.html .

Related Material