Researchgroups in The Netherlands

Multi-agent and adaptive computing, CWI

The Multi-agent and Adaptive Computation, group headed by Prof Dr Ir Han La Poutré, focuses on the design of adaptive computational techniques for decision making in decentralized and complex environments. This comprises (1) computational intelligence techniques, like evolutionary algorithms, adaptive algorithms, graphical models, and hybrid heuristics; and (2) decentralized systems, like multi-agent systems, sensor networks, and non-cooperative games.

> http://www.cwi.nl/research-groups/multi-agent-and-adaptive-computation

Decentralized decision making often requires the local decision makers, such as agents, intelligent sensors, or services, to be adaptive to their environment (computational learning), while the appropriate rules to which these decision makers should commit are important design issues in order to get the desired collective behavior (mechanism design) and results.

Key problem areas concern economic optimization problems (management and logistic services), collaborative discovery of semantics in distributed sensor networks, and software services (logistic services and market-based systems). In these problem areas, decentralized and adaptive decision making becomes more and more important and fits the actual practices and requirements in reality.

 
The activities of the group address e.g. the fundamental development of the underlying CI techniques; the design of adaptive agent strategies in economic games (like negotiation, auctions, and other mechanisms); the design of economic games, agents, and simulation systems (as a part of computational science); and the development of generic models for application areas that require such solutions.

Application domains include management of electricity networks, smart environments, health care logistics, service markets, the Internet of Things, transportation logistics, and market simulation.

 
Prof Dr Ir Han La Poutré
Dr Erik Pauwels
Dr Peter Bosman