Bio
Sanghyun Choo is a postdoctoral research associate in the Scalable Biomedical Modeling group, part of the Computational Sciences and Engineering Division at Oak Ridge National Laboratory (ORNL).
Prior to joining the ORNL, Sanghyun's research focused on model-agnostic/-specific post hoc methods in eXplainable Artificial Intelligence (XAI) for Brain-Computer Interface (BCI), representation learning for various applications (e.g., signal, image, text), regularization methods (e.g., data augmentation, adaptive batch size) to improve the generalization performance of Deep Learning (DL), uncertainty-aware Interactive Reinforcement Learning (IRL) to reduce sample complexity, performance improvement of Machine Learning (ML) models, brain network analysis to find causal relationships among brain components, etc.
Currently, Sanghyun's research focuses on developing privacy-preserving ML and DL pipelines with supercomputing and HPC in Healthcare. His research interests also include federated learning, differential privacy, and multimodal learning.
Education
Ph.D. in Industrial & Systems Engineering at North Carolina State University