Comput Methods Biomech Biomed Engin. 2026 Jan 10:1-18. doi: 10.1080/10255842.2025.2612536. Online ahead of print.
ABSTRACT
To mitigate the limitations of Light Gradient Boosting Machine (LightGBM) in processing heterogeneous cardiovascular disease (CVD) data, a Hierarchical Quantum Ensemble Model (HQEM) is proposed. This architecture deploys a Quantum Neural Network (QNN) and eXtreme Gradient Boosting (XGBoost) as parallel base classifiers to capture non-linear quantum patterns and sequential gradient trends. The resulting ensemble outputs enrich the feature space for a LightGBM meta-classifier. Validation across integrated datasets yielded 97% accuracy and a 98% Area Under the Curve (AUC), demonstrating the model's superior efficacy in handling complex feature distributions for robust CVD classification.
PMID:41517992 | DOI:10.1080/10255842.2025.2612536