RULES MINING FOR THE DIAGNOSIS OF BREAST ADENOCARCINOMA USING ROUGH SET THEORY

In this paper a Rough Set Theory based approach is used to mine concise rules from a huge dataset of patients treated for Breast Adenocarcinoma, one of the prominent type of Adenocarcinoma among women. This paper works on a sample dataset containing data of six different classes, organized as objects and attributes. The dataset contains lot of uninformative attributes which has to be filtered out. Various algorithms are applied based on Rough Set Theory to first eliminate noise and redundancy from the dataset and then generate rules .Different set of rules are generated for different reducts obtained by the application of different algorithms. These rules are studied thoroughly to understand the behavior of the patients’ genes diagnosed with Breast Adenocarcinoma. Most important rules are singled out, using metric values, to get a good knowledge about the reasons behind the decision classes.