Federated learning framework with hybrid feature selection for heart disease prediction
Year: 2026
Type: article
Abstract: Since cardiovascular diseases are a leading cause of death worldwide, accurate and privacy-preserving prediction methods are essential. Traditional machine learning models often rely on centralized data, which raises issues with data security, privacy, and heterogeneous data distribution among healthcare organizations. This study proposes a federated learning framework with hybrid feature selection (FL-HFS) for heart disease prediction to overcome these challenges. The framework assures data confidentiality by enabling collaborative model training without exchanging raw patient data. In order to determine important clinical features and improve model efficiency, a hybrid feature selection methodology that combines variance thresholding, mutual information, and L1 regularization is implemented. The model is evaluated on a diverse Dataset under a federated learning framework. Experimental results show that the proposed method outperforms baseline federated learning models with an accuracy of 88.3% and an F1-score of 91%. The results show that the proposed framework provides a robust, scalable, and privacy-aware solution for distributed healthcare applications. Clinical trial number : Not applicable.
Source: Discover Computing
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $1,520
OpenAlex ID: W7225660470