Intelligent decision making in medical data using association rules mining and fuzzy analytic hierarchy process

Abstract

The confidence and support value are two measures that playa signification role to make Association rules high important and widely acceptable. In parallel, the length of rule and the presence of more significant features in rule increase its acceptability. The selection of some high important rules based on these measures is a difficult task. Analytic Hierarchy Process (AHP) provides decision matrix with weight value of each measures (factors), which helps in ranking the rules based on measures. But, AHP is not capable to take perfect decision in the case where the rules have some uncertainty or fuzziness, especially in Medical Association rules. The proposed work discusses the fuzzy rule base Analytic Hierarchy process to evaluate the relative (importance) weight of different measures in order to choose the perfect rule. Here, liver disorder medical data is used to generate Association rules and then fuzzy AHP based method is applied to make comparison matrix and different rules are compared using TFNs.