IEICE Technical Committee Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2024-01-17 14:30
Extrinsicaly Rewarded Soft Q Imitation Learning with Discriminator
Ryoma Furuyama, Daiki Kuyoshi, Yamane Satoshi (Kanazawa Univ.) MSS2023-55 SS2023-34
Abstract (in Japanese) (See Japanese page) 
(in English) Imitation learning is often used in addition to reinforcement learning in environments where reward design is difficult or where the reward is sparse, but it is difficult to be able to imitate well in unknown states from a small amount of expert data and sampling data. Supervised learning methods such as Behavioral Cloning(BC) do not require sampling data, but usually suffer from distribution shift. The methods based on reinforcement learning, such as inverse reinforcement learning and Generative Adversarial imitation learning (GAIL), can learn from only a few expert data. However, they often need to interact with the environment. Soft Q imitation learning (SQIL) addressed the problems, and it was shown that it could learn efficiently by combining Behavioral Cloning and soft Q-learning with constant rewards. In order to make this algorithm more robust to distribution shift, we propose more efficient and robust algorithm by adding to this method a reward function based on adversarial inverse reinforcement learning that rewards the agent for performing actions in status similar to the demo. We call this algorithm Discriminator Soft Q Imitation Learning (DSQIL) . We evaluated it on MuJoCo environments.
Keyword (in Japanese) (See Japanese page) 
(in English) Artificial Intelligence / Machine Learning / Deep Reinforcement Learning / Imitation Learning / / / /  
Reference Info. IEICE Tech. Rep., vol. 123, no. 334, MSS2023-55, pp. 19-24, Jan. 2024.
Paper # MSS2023-55 
Date of Issue 2024-01-10 (MSS, SS) 
ISSN Online edition: ISSN 2432-6380
Copyright
and
reproduction
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034)
Download PDF MSS2023-55 SS2023-34

Conference Information
Committee SS MSS  
Conference Date 2024-01-17 - 2024-01-18 
Place (in Japanese) (See Japanese page) 
Place (in English)  
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To MSS 
Conference Code 2024-01-SS-MSS 
Language English 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Extrinsicaly Rewarded Soft Q Imitation Learning with Discriminator 
Sub Title (in English)  
Keyword(1) Artificial Intelligence  
Keyword(2) Machine Learning  
Keyword(3) Deep Reinforcement Learning  
Keyword(4) Imitation Learning  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Ryoma Furuyama  
1st Author's Affiliation Kanazawa University (Kanazawa Univ.)
2nd Author's Name Daiki Kuyoshi  
2nd Author's Affiliation Kanazawa University (Kanazawa Univ.)
3rd Author's Name Yamane Satoshi  
3rd Author's Affiliation Kanazawa University (Kanazawa Univ.)
4th Author's Name  
4th Author's Affiliation ()
5th Author's Name  
5th Author's Affiliation ()
6th Author's Name  
6th Author's Affiliation ()
7th Author's Name  
7th Author's Affiliation ()
8th Author's Name  
8th Author's Affiliation ()
9th Author's Name  
9th Author's Affiliation ()
10th Author's Name  
10th Author's Affiliation ()
11th Author's Name  
11th Author's Affiliation ()
12th Author's Name  
12th Author's Affiliation ()
13th Author's Name  
13th Author's Affiliation ()
14th Author's Name  
14th Author's Affiliation ()
15th Author's Name  
15th Author's Affiliation ()
16th Author's Name  
16th Author's Affiliation ()
17th Author's Name  
17th Author's Affiliation ()
18th Author's Name  
18th Author's Affiliation ()
19th Author's Name  
19th Author's Affiliation ()
20th Author's Name  
20th Author's Affiliation ()
Speaker Author-1 
Date Time 2024-01-17 14:30:00 
Presentation Time 25 minutes 
Registration for MSS 
Paper # MSS2023-55, SS2023-34 
Volume (vol) vol.123 
Number (no) no.334(MSS), no.335(SS) 
Page pp.19-24 
#Pages
Date of Issue 2024-01-10 (MSS, SS) 


[Return to Top Page]

[Return to IEICE Web Page]


The Institute of Electronics, Information and Communication Engineers (IEICE), Japan