Graphics enhanced version of this site

Sachiyo Arai
Postdoctoral Fellow

No longer a member of RI.

Jump to: Research interests | Keywords | Additional Interests and Responsibilities | Labs & Groups | Projects | Publications

Cooperative and coordinated behavior that are observable in animal societies are very attractive and have been the motivation for my research for recent years.

I am investigating two key areas in the multi-agent domain. One is to throw light on the mechanism of generating cooperation. The other is to realize the cooperative behavior in practical systems by means of machine learning frameworks to integrate the top-down knowledge (symbolic and structured) and bottom-up knowledge (acquired through reinforcement learning and evolution).

I have proposed experience-based reinforcement learning algorithm which is very effective to emerge adaptive behaviors in the dynamic domain where includes multiple learning-entities. This algorithm works well even in un uncertain domain where agents can not have perfect information around the world.

education, game theory, machine learning, and planning

Education and teaching are very important in making children, robots and other kind of creatures do the right thing.

• Intelligent Software Agents - Here, researchers explore multi-agent planning and scheduling, multi-agent learning, multi-agent negotiation, and decision support for human teams

• Adaptive Interoperability of Multiple Heterogeneous Agents - To enable requester agents to find provider agents in environments with heterogeneous and dynamic information services, we have developed middle agents which serve to increase agent interoperability.
• Pleiades - A project aimed at developing methods for automating negotiation among agents and implementing effective machine learning techniques.

Note: This list may not be comprehensive. It contains only those publications in the RI publications database. Entries are listed in reverse chronological order.

• Multi-agent reinforcement learning for planning and scheduling multiple goals
  S. Arai, K. Sycara, and T. Payne
  Proceedings of the Fourth International Conference on MultiAgent Systems, July, 2000, pp. 359 - 360. [Abstract]
  Download: pdf [172 KB] copyrighted

• Experience-based Reinforcement Learning to Acquire Effective Behavior in a Multiagent Domain
  S. Arai, K. Sycara, and T. Payne
  The Sixth Pacific Rim International Conference on Artificial Intelligence (PRICAI 2000), Springer-Verlag, 2000, pp. 125-135. [Abstract]
  Download: pdf [270 KB] copyrighted