Connectionist decision support systems for multiple criteria decision making

The authors present a general framework for connectionist decision support systems for solving deterministic multiple criteria decision problems. The decision support systems are driven by preferential data and based on prescriptive decision models, with artificial neural networks in their core. They are able to interact with decision makers and construct, verify, and validate themselves autonomously from available preferential data. As time elapses, the decision support systems are also capable of refining their internal representations accordingly to adapt to potential changes of decision makers' preference and decision environments over time.<<ETX>>