A Hybrid Inference Framework for Model Selection

A hybrid inference framework is presented in which the aim is to select decision models through natural language in knowledge-based decision support system. In situations in which the rules that determine a system are uncertain and incomplete, the knowledge elicitation, maintenance and update of the rule-based system can be the problematic tasks. In such a situation, it has been found that a hybrid case-based and rule-based reasoning framework can provide a more effective and accurate means of performing such selection than other statistical methods and models. The hybrid framework results in capturing knowledge in context by storage of the case that prompted a new rule to be added, which aims at the unwanted side effects associated with typical rule reasoning system. This framework has been used to determine the inventory models in the decision support system. The results obtained from the experiment are presented with the both efficiency-improving and accuracy-improving