Prediction of Heart Disease by Feature Selection Technique using Particle Swarm Optimization based on Logistic Regression

Abstract

Heart disease is one of the main causes of illness and mortality in the modern world. in the world. Most of the deaths are occurring due to heart failure or some cardiovascular diseases. Therefore, it becomes crucial to identify abnormal heart behavior at an early stage. The prediction of disease at an early stage can lead to effective prevention and treatment. In this paper, we propose a novel approach for predicting heart disease by integrating feature selection techniques with Particle Swarm Optimization (PSO) and Logistic Regression (LR). A feature selection method based on PSO is used to increase prediction accuracy. PSO maximizes the predictive performance by reducing complexity to find the ideal combination of features for the feature subset iteratively will help the prediction model's performance. . Our results show that the proposed approach of using PSO to optimize the feature selection process of Logistic Regression for predicting heart disease is a promising idea that could increase the prediction of heart disease accuracy. The technique might enable medical practitioners to detect cardiac illness earlier and make better educated judgments that could result in earlier interventions and therapies that could save lives.