Hyperband-tuned XGBoost leads atrial fibrillation detection study
Original title: Enhancing Atrial Fibrillation Detection Through Hyperparameter Optimization
On PhysioNet ECG data the best model reached a peak F1 score of 96.95%.
The study compares Decision Tree, Random Forest, and XGBoost with recurrent, deep, and Siamese neural networks for atrial fibrillation detection. Four tuners—Genetic Algorithm, Hyperband, Optuna, and SMAC3—are applied under inter-patient and intra-patient splits that include class imbalance. Hyperband-optimized XGBoost ranks first on PhysioNet electrocardiogram datasets.
- Models include Decision Tree, Random Forest, XGBoost, RNNs, DNNs, and Siamese networks.
- Tuners are Genetic Algorithm, Hyperband, Optuna, and SMAC3.
- Splits are inter-patient and intra-patient, reflecting real-world AF imbalance.
- Peak reported F1 score is 96.95% for Hyperband-optimized XGBoost.
From the original article
1 min read at arxiv.orgThis study explores the role of hyperparameter optimization (HPO) in enhancing the detection of Atrial Fibrillation (AF) using both machine learning and deep learning models. We evaluated a range of standard classifiers - Decision Tree, Random Forest, and XGBoost classifiers, alongside deep learning architectures including Recurrent Neural Networks, Deep Neural Networks, and Siamese Neural Networks.
By Sara Gjorgjieva · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.