A Thrice Filtered Information Energy Optimization Based Feature Selection (TFIE-OFS) Method for Heart Disease Prediction

Abstract

Heart disease remains a leading cause of mortality worldwide, necessitating effective classification and prediction methods to enhance early detection and intervention. This study proposes a novel Thrice Filtered Information Energy Optimization based Feature Selection (TFIE-OFS) method, which integrates Symmetrical Uncertainty, Information Gain, and Chi-Square Analysis to systematically filter and prioritize features from heart disease datasets. By employing Particle Swarm Optimization (PSO), the TFIE-OFS method optimizes feature subsets, ensuring the selection of the most informative variables while minimizing redundancy. The efficacy of the proposed method is evaluated through comprehensive experiments on benchmark heart disease datasets, where it demonstrates superior classification performance compared to existing feature selection techniques. The results indicate that TFIE-OFS significantly enhances predictive accuracy and model interpretability, providing a robust framework for heart disease classification and prediction. This innovative approach not only contributes to the field of medical data analytics but also holds potential for improving clinical decision-making in cardiology.