Unexpected Health Issues Prediction In Medical Data Using Apriori Rare Based Outlier Detection Method

People affected by hazardous diseases arise in recent years. The growth of the fatal diseases is at an alarming rate. Initial detection can be helpful to get relieved from the disease. Thus the techniques for the early stage identification is much required. Existing classification algorithms employed tend to distinguish normal and disease affected people and the limitation involves no beforehand awareness about the occurrence of the disease. Outlier detection has vital application in the medical research domain for alarming the patients priorly. Exploring outliers from the medical domain gives useful and effective information to the domain experts. Hence, we propose an Apriori-rare based Associative Classification technique to predict unexpected issues in the medical domain, which is accomplished by the discovering the outliers. Healthcare datasets like Mammography mass data, Chronic kidney disease and Heart disease from UCI repository are considered for diagnosis and prognosis. The built classifier by the suggested method, enhances the accuracy and prediction rate of the diseases by detecting the presence of outliers. The outcomes are compared with a Classification Based on Association rules (CBA) algorithm in R Tool. The proposed Apriori rare base associative classification algorithm excels than the existing CBA algorithm by 15% more accurate.