Machine Learning Based Predictive Analysis of Diseases in Health Care

Prediction and classification of diseases in the healthcare sector are one of the major challenges for treating patients on time. Cardiovascular disease, in particular, is a major threat that raises the global death rate. This global death rate can be reduced by influencing the hybridization of medical and information technology together in the health care sector. The vital sign parameters are continuously measured and can be used to predict and classify cardiovascular disease and alert the healthcare practitioner to treat the patient. The existing research used various machine learning algorithms and conducted comparison analyses. This work proposes a machine-learning-based model for the prediction and classification of cardiovascular diseases using a vital sign dataset. Based on the prediction outcome, the caretaker or health care experts can understand the health status of the patient. The dataset used in this work has been taken from the Queensland University repository; the “10 minutes and 20 minutes” data have been considered. The multivariate regression technique has been used for prediction, and the Random Forest algorithm, Support Vector Machine (SVM), and Multilayer Perceptron Network have been used for classification. The prediction model predicts with 96.4% accuracy, and the Random Forest algorithm classifies the patient data as “risky” or “not risky” with 98.14% accuracy, which is better than SVM and Multilayer Perceptron Network.