Projects per year
Abstract
The ability of an AI agent to assist other agents, such as humans, is an important and challenging goal, which requires the assisting agent to reason about the behavior and infer the goals of the assisted agent. Training such an ability by using reinforcement learning usually requires large amounts of online training, which is difficult and costly. On the other hand, offline data about the behavior of the assisted agent might be available, but is non-trivial to take advantage of by methods such as offline reinforcement learning. We introduce methods where the capability to create a representation of the behavior is first pre-trained with offline data, after which only a small amount of interaction data is needed to learn an assisting policy. We test the setting in a gridworld where the helper agent has the capability to manipulate the environment of the assisted artificial agents, and introduce three different scenarios where the assistance considerably improves the performance of the assisted agents.
Original language | English |
---|---|
Title of host publication | Artificial Neural Networks and Machine Learning – ICANN 2021 - 30th International Conference on Artificial Neural Networks, Proceedings |
Editors | Igor Farkaš, Paolo Masulli, Sebastian Otte, Stefan Wermter |
Publisher | Springer |
Pages | 519-530 |
Number of pages | 12 |
ISBN (Print) | 978-3-030-86379-1 |
DOIs | |
Publication status | Published - 2021 |
MoE publication type | A4 Conference publication |
Event | International Conference on Artificial Neural Networks - Virtual, Online Duration: 14 Sept 2021 → 17 Sept 2021 Conference number: 30 |
Publication series
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Volume | 12894 LNCS |
ISSN (Print) | 0302-9743 |
ISSN (Electronic) | 1611-3349 |
Conference
Conference | International Conference on Artificial Neural Networks |
---|---|
Abbreviated title | ICANN |
City | Virtual, Online |
Period | 14/09/2021 → 17/09/2021 |
Keywords
- Cooperative AI
- Deep reinforcement learning
- Helper agent
- Meta-learning
- Modelling other agents
Fingerprint
Dive into the research topics of 'Learning to Assist Agents by Observing Them'. Together they form a unique fingerprint.Projects
- 2 Finished
-
Interactive machine learning from multiple biodata sources
Kaski, S. (Principal investigator), Hämäläinen, A. (Project Member), Gadd, C. (Project Member), Hegde, P. (Project Member), Shen, Z. (Project Member), Siren, J. (Project Member), Trinh, T. (Project Member), Jain, A. (Project Member) & Jälkö, J. (Project Member)
01/01/2019 → 31/08/2021
Project: Academy of Finland: Other research funding
-
Interactive machine learning from multiple biodata sources
Kaski, S. (Principal investigator) & Filstroff, L. (Project Member)
01/01/2016 → 31/08/2021
Project: Academy of Finland: Other research funding