An Enhanced Probabilistic Elastic Net Regression Model (EPERM) for Heart Disease Prediction
V. Tamil Selvi, T. Ganesh Kumar, J.B. Shajilin Loret, K. Sampath Kumar, Anitha Julian, Priti Rishi
Abstract
Heart attack prediction is a significant contributor to worldwide morbidity. Cardiovascular illness is a critical component of clinical data analysis prediction. A precise diagnosis of cardiac disease can save lives, yet an inaccurate diagnosis can lead to death. To predict heart disease prior, an Enhanced Probabilistic Elastic Net Regression model (EPERM) is presented, which is utilized to classify cardiac disease, and every feature selection is based on correlation with the goal value. The Attribute Ranking Method was used to select and organize highly positive associated attributes based on their weight. Finally, conditional probability is utilized to estimate the likelihood of cardiac disease in advance. This study makes use of one of the largest disease-related databases. Regarding projected illness progression, the suggested algorithm outperforms the most recent techniques. The findings also demonstrate how a patient’s disease sequence may alter their risk management in the future. Our prediction model, based on Elastic Net Regression and data analysis, discovered several traits that clinical knowledge had previously overlooked. This article investigates clinical diagnostic records from the past decade to identify probable heart cancer symptoms and comorbidities. It also provides early disease prediction, allowing a greater number of patients to receive an early diagnosis. The suggested model accurately predicts evidence of a person’s heart illness using Probability-based Elastic Net Regression. The results show that the technique is a reliable and consistent prediction model.