Heart Disease Prediction Using Feature Selection and Classification Models

• In Event: HEALTHCARE MANAGEMENT: Machine Learning in Healthcare

Abstract

Heart disease is the leading cause of mortality worldwide, and hospitals provide a plethora of medical information about it. In this research work, we conducted an operational evaluation of models constructed using ten feature selection approaches to improve heart disease accuracy and found that the Backward Feature Selection (BFS) procedure had the highest levels of classification accuracy (89%), precision (91%), sensitivity (81%), and f-measure (86%), compared to the previous studies. This study suggests that BFS is a promising feature selection technique for improving the accuracy of classification models for heart disease.

Authors

• Dinesh K Sharma, University of Maryland Eastern Shore

• Tarun Dhar Diwan, Atal Bihari Vajpayee University

• Ajay Tiwari, Atal Bihari Vajpayee University