Ameliorating Heart Diseases Prediction using Machine Learning Technique for Optimal Solution

Abstract

In the era of lacking physical fitness, folks in society are facing vital health complications which can be due to a variety of reasons such as pollution, work pressure, food, and sleeping patterns, etc. One such critical health issue which can significantly affect the regular physical activity of patients is heart disease, emulating early predictions can protect lives by inculcating proposed work and improved methodology for early heart disease prediction. In this paper, an improved methodology for heart disease prediction is proposed using a logistic regression classifier, by tuning hyperparameter using a grid-based solver with tenfold cross-validation. The dataset used in the work is driven from the UCI Machine learning repository to evaluate the efficacy of the proposed model. Medical researchers or doctors can evaluate the model’s accuracy and gets its performance. The enhanced performance rate of 90.16% has been shown in the experimental results which portrays it outclasses many prevailing models. Using this upgraded model can truly reduce the mortality rate of heart disease patients. In addition to the proposed model to make it user-friendly, a user interface has been designed where symptoms can be given as input and receives prediction as output.