Prognosis of Supervised Machine Learning Algorithms in Healthcare Sector

Medical care is a fundamental liberty. The conquering application of Machine Learning (ML) in this computerized world is noticeable. With the increase in medical data, ML is penetrating in medical care industry resulting in the integration of Machine Learning algorithms and knowledge of medical personnel and designing of prognostic models which can help doctors and patients to analyze risks of any health compilation. Researchers from health domain are exploring ML algorithms to reach out to some useful conclusions. With this paper we aim to help the researchers to understand the efficiency of available ML algorithms on medical datasets, thus helping them to decide which one to choose from the existing methods. This paper implements 5 Supervised Machine Learning algorithms on four different datasets from health domain on Heart Disease, Diabetes, Dermatology, and Breast Cancer. Results of each of the implemented ML algorithms are compared in terms of prediction accuracy and AUC value on medical datasets. Implementation results suggests that Logistic Regression and Random Forest have shown better results with almost all the datasets used for experiment purpose with accuracy (85%-88%) and AUC value (0.89-0.92). The yield of this paper will add to a better understanding of the use of Machine Learning in the Medical Domain.