Evolutionary selection of model for time constrained decision problems: A GA approach

The purpose of this paper is to develop a model selection method for time constrained decision problems. Usually there are many models in a decision support system (DSS), thus the model management system (MMS) is an important component of a DSS. In this paper, we study a DSS that offers competitive models with different levels of solution accuracy. The goodness of a model will be judged by its payoff value that involves the model solution accuracy and information costs. A Genetic Algorithm (GA) is used to find the maximum payoff value of a model, and the model with the best payoff value is selected for the target problem. We also propose a model selection system that is easy to implement with the help of web services technology.