DefinePK

Abstract

In recent times, heart disease has emerged as a significant global health concern. The utilization of machine learning, deep learning, and other artificial intelligence (AI) tools to aid in medical diagnostics is steadily gaining traction.This work presents a unique LSKR soft voting ensemble model with Firefly optimization (LSKR-SVE(FO)), which consists of four different learners, to improve the prediction accuracy of heart disease. Here, PrincipalComponent Analysis and Lasso method have been utilized in feature extraction and feature selection respectively for enhancing the prediction accuracy of LSKR-SVE(FO). The weight value of SVE and overfitting problem can be minimized by utilizing Firefly optimization algorithm. This study investigates efficient heart disease diagnosis using the heart disease dataset from the UCI Machine Repository. Proposed LSKR-SVE(FO) achieved the highest performance (99.3% accuracy, 98.45% precision, 96.2% Recall), followed by SVM-AHP (96.3% accuracy, 98.5% precision, 88.3% recall). These findings suggest that our optimized algorithm offers an effective healthcare monitoring system for early heart disease prediction