Data-Driven Insights: A Genetic Algorithm Feature Optimization Approach to Heart Disease Prediction

Heart disease (HD) encompasses various conditions affecting the cardiovascular system's well-being. In the current global context, early detection of heart disease is exceptionally challenging, and delaying its identification could result in severe consequences. This process typically involves time-consuming and potentially subjective medical assessments. To address this challenge, machine learning (ML) techniques have been harnessed for autonomous data analysis. This research introduces a significant element of using Genetic Algorithm (GA) within metaheuristic optimization techniques for feature optimization. The study assesses a dataset from the UCI using standard models and subsequently identifies gaps in existing literature, presenting its findings in the comparative analysis section. The author's analysis indicates that GA exhibits superior performance with an accuracy of 96.72% in Elastic Net, Logistic Regression, and Gradient Boosting algorithms. The experimental findings illustrate that the suggested method excels in comparison to current approaches, especially in the realm of feature selection (FS) techniques. This survey holds value for researchers interested in employing ML and data mining (DM) strategies for HD prediction.