A Computational Intelligence Technique for Effective and Early Diabetes Detection using Rough Set Theory

Abstract

Huge amount of medical databases requires sophisticated techniques for storing, accessing, analysis and efficient use of stored acquaintance, knowledge and information. In early days intelligent methods like neural networks, support vector machines, decision trees, fuzzy sets and expert systems are widely used in the medical fields. In recent years rough set theory is used to identify the data associations, reduction of data, data classification and for obtaining association rules form the mined databases. In this research contribution we proposed a method for generating association classification rules for the classification of Pima Indian Diabetes (PID) data set taken from UCIML repository. We obtained promising results with this method on the PID data set. General Terms Expert systems, fuzzy sets, decision trees