Prediction of Heart Disease Using Genetic Algorithm for Selection of Optimal Reduced Set of Attributes

The healthcare industry collects huge amounts of healthcare data which is not feasible to handle manually. Due to advancements in technology, today, many hospitals use information systems to store and manage data. These large amounts of data are very important in the field of data mining to extract useful information and generate relationships amongst the attributes. Thus, data mining is used to develop a mechanism to predict risk of heart disease. With the tremendously growing population, the doctors and experts available are not in proportion with the population. Also, symptoms of heart disease may not be significant and thus, may often be neglected. So we propose an Intelligent Heart Disease Decision Support System to help the doctors reach out to those people who are deprived of these medical services. In general, it can serve as a training tool to train nurses and medical students to diagnose patients having risk of heart disease. In this paper we have discussed two modeling techniques: Naïve Bayes’ Rule and Genetic Algorithm which predict the risk level of heart disease. In Genetic Algorithm, optimal reduced set of attributes are found using Genetic Search method. The 13 attributes in the original list have been reduced to 6. In Naïve Bayes’ technique, a historical heart disease database is used to generate relationships amongst the attributes using the concepts of conditional probability.