An artificial immune system as a multi-agent decision support system

The paper proposes a general framework for building an intelligent decision support system based on immunological principles. From the computational viewpoint, the natural immune system is a highly decentralized information processing system that can generate appropriate responses in order to provide maximum protection against foreign antigens. The paper examines various recognition and response mechanisms of the immune system to develop a massively parallel adaptive decision support system. Such an integrated system can also be viewed as a multiagent system where the functionalities and the capabilities of different types of agents vary. Moreover, the agents may move and interact freely in the environment with other agents. They can mutually recognize each others activities and can produce specific response based on predefined decision strategies. The purpose of this research is to demonstrate that such a collaborative multi-agent system can enhance the decision making process, and can solve complex tasks more precisely and efficiently. A prototype system is currently under implementation in an object-oriented software paradigm with visualization tools.