Comparative Analysis of Data Mining Algorithms on EHR of Rheumatoid Arthritis of Multiple Systems of Medicine

Abstract

Data mining techniques are applied usually to uncover concealed knowledge from massive data stacked up in databases. One of the potential fields of Data mining application is healthcare systems in which the increasingly large amount of data are populated in the databases. Such populated databases needs the application of suitable data mining techniques to extract the knowledge patterns which are vital decision making as well as care taking systems. In the field of healthcare enormous amount of data is generated and populated in databases. These databases are vital for knowledge extraction and its uses for futuristic betterment of health of populace. The Electronic Health Record (EHR) database for a disease of Rheumatoid Arthritis is considered in the research work. It includes the data from multiple systems of medicine which include Ayurvedic system of medicine and Allopathic system of medicine. The classification algorithms - BayesNet, Naïve Bayes, ZeroR, JRip, OneR and PART are implemented on EHR of Rheumatoid Arthritis. Results are obtained for 100, 500 and 1000 instances of EHR to encompass a comparative approach for analytics.