HEALTHCARE DATA ANALYSIS UTILIZING AN ENSEMBLE FEATURE SELECTION TECHNIQUE

Abstract

The analysis of healthcare data is critical for enhancing clinical decision-making and improving patient outcomes. However, the high dimensionality and presence of noisy, irrelevant features in structured healthcare datasets present significant challenges. This paper proposes an ensemble feature selection technique specifically designed to handle these challenges by selecting the most relevant features for classification tasks. The proposed Competitive Ensemble Feature Selection Model (CEFSM) integrates multiple feature selection methods, each contributing to the identification of optimal features. The effectiveness of the CEFSM is validated using various healthcare datasets, demonstrating its ability to improve classification accuracy and reduce computational complexity. The paper also explores the application of this feature selection technique within a Competitive Ensemble Classification Model (CECMSDML) to further enhance the performance of healthcare data analysis.