Indian Journal of Science and Technology

Abstract

Objective: The industry of healthcare contains large information, which is difficult to process by manual methods. The large data are too valuable for extracting information and forming relationship from data mining area. Analysis: The experts and experienced doctors are not also available against the large population; sometimes symptoms are also being neglected. The existing system in the medical field is not able to extract all the information and knowledge from the heart disease database. Complex query for the healthcare practitioner to analyze the heart disease is still a challenging task. Finding: This paper is presenting a novel technique like K-Means, Weighted Associative Classifier (WAC) and Prediction Tree C5.0 for analyzing the heart disease and to sort out the existing issues. The K-Means is being used for unsupervised learning cluster within WAC. Initial centroid is being selected by the K-Means, which allow the classifier to extract the record and make a prediction for analyzing the disease with C5.0 prediction tree. The combined technology of K-Means, WAC and Prediction Tree C5.0 will provide a better, integrated, and accurate result over the heart disease prediction. The projected tool is MapReduce with HiveDatabase in Hadoop open source framework. Hadoop is perfectly compatible for big data projects. Improvements: The approaches for developing an Intelligent System for Heart Disease Prediction with big data mining will be very advantageous by automation of the proposed system K-Means, Prediction Tree C5.0, Weighted Association Classifier (WAC).

Keywords: Hive Database, K-Means Clustering, MapReduce, Prediction Tree C5.0, Weighted Association Classifier, Weighted Support and Confidence