Deep reward shaping from demonstrations.

Hussein, Ahmed; Elyan, Eyad; Gaber, Mohamed Medhat; Jayne, Chrisina

Abstract

Deep reinforcement learning is rapidly gaining attention due to recent successes in a variety of problems. The combination of deep learning and reinforcement learning allows for a generic learning process that does not consider specific knowledge of the task. However, learning from scratch becomes more difficult when tasks involve long trajectories with delayed rewards. The chances of finding the rewards using trial and error become much smaller compared to tasks where the agent continuously interacts with the environment. This is the case in many real life applications which poses a limitation to current methods. In this paper we propose a novel method for combining learning from demonstrations and experience to expedite and improve deep reinforcement learning. Demonstrations from a teacher are used to shape a potential reward function by training a deep supervised convolutional neural network. The shaped function is added to the reward function used in deep-Q-learning (DQN) to perform off-policy training through trial and error. The proposed method is demonstrated on navigation tasks that are learned from raw pixels without utilizing any knowledge of the problem. Navigation tasks represent a typical AI problem that is relevant to many real applications and where only delayed rewards (usually terminal) are available to the agent. The results show that using the proposed shaped rewards significantly improves the performance of the agent over standard DQN. This improvement is more pronounced the sparser the rewards are.

Citation

HUSSEIN, A., ELYAN, E., GABER, M.M. and JAYNE, C. 2017. Deep reward shaping from demonstrations. In Proceedings of the 2017 International joint conference on neural networks (IJCNN 2017), 14-19 May 2017, Anchorage, USA. Piscataway: IEEE [online], article number 7965896, pages 510-517. Available from: https://doi.org/10.1109/IJCNN.2017.7965896

• Conference Name: 2017 International joint conference on neural networks (IJCNN 2017)
• Conference Location: Anchorage, USA
• Start Date: May 14, 2017
• End Date: May 19, 2017
• Acceptance Date: Feb 3, 2017
• Online Publication Date: May 14, 2017
• Publication Date: Jul 3, 2017
• Deposit Date: Mar 9, 2017
• Publicly Available Date: Mar 9, 2017
• Print ISSN: 2161-4393
• Electronic ISSN: 2161-4407
• Publisher: Institute of Electrical and Electronics Engineers (IEEE)
• Article Number: 7965896
• Pages: 510-517
• Series ISSN: 2161-4407
• ISBN: 9781509061815
• DOI: https://doi.org/10.1109/IJCNN.2017.7965896
• Keywords: Deep reinforcement; Reinforcement learning; Generic learning process; DeepQlearning (DQN)
• Public URL: http://hdl.handle.net/10059/2196