Solving Complex Decision-Making Problems Through Agent-Matrices Cooperation
Hao Lan Zhang,Jiming Liu,Yong Tang,Chaoyi Pang
DOI: https://doi.org/10.1007/978-3-642-41154-0_37
2013-01-01
Abstract:Making decisions in a dynamic environment is complicated and indefinite because of various unpredictable factors and large volumes of information. It has become a key issue for modern organizations to have efficient systems and tools to support rational decision-making in problem-solving processes, particularly, for problems with a lack of pre-defined structure. This paper provides a novel agent-based framework for Decision Support Systems (DSS) based on the analysis of the major problems in existing DSS. The proposed multi-agent-based framework, namely the Agent-Matrices framework, provides an open, efficient and flexible architecture for DSSs. The Agent-Matrices framework overcomes the compatibility and connectivity problems in most traditional DSS applications. This framework utilises Matrices to allow agents to acquire information, self-upgrade, perform tasks, travel to other Matrices, and be reused. This paper primarily introduces the methods used for coordinating the Matrices and agents to solve a complex problem.