Publications by authors named "Sun Cheol Heo"
Article Synopsis
- Federated learning in healthcare enables collaboration on model training using distributed data while maintaining privacy; however, traditional methods struggle to utilize unique institutional data features.*
-
- A new method called personalized progressive federated learning (PPFL) was proposed, which considers client-specific features and showed superior performance in in-hospital mortality prediction, with an accuracy of 0.941 and AUROC of 0.948.*
-
- PPFL not only outperformed conventional federated models but also retained strong performance with cancer data, identifying key features linked to mortality for different institutions.*
