Hello 👋

I am a Data Scientist in Professor Sophia Wang’s research group, Ophthalmic Informatics and Artificial Intelligence Group, at the Stanford School of Medicine. The group broadly focuses on using a variety of data modalities (i.e., EHR, clinical notes, imaging) to build predictive models for glaucoma. Currently, my work focuses on building predictive models for glaucoma surgery progression using a multi-center dataset, SOURCE. Some challenges I attempt to solve are data irregularity, sparsity, heterogeneity, and model opacity.

Previously at Columbia University as a graduate student, I worked closely with Professor Ansaf Salleb-Aouissi in her lab PRAISE. My work there focused on building predictive models for spinal surgery outcomes. ​​My theoretical work was heavily based on knowledge distillation and privileged information due to the rarity of the surgery, thus, the data as well. I proposed a variant of XGBoost, XGBoost+, that uses privileged information (data available during training but not at inference), to increase the model’s predictive power. I show empirically that this model performs better than its parent, XGBoost, and other tree-based models.

My research interests lie in the intersection of machine learning and healthcare. I am interested in solving challenges such as representation learning, explainability, and multi-model learning. These are key challenges that must be solved before bringing our research to a clinical setting.