I'm a Statistics MSc student at ETH Zurich with a passion for machine learning theory and applications in healthcare and climate science. My interests center around interpretability, probabilistic machine learning, and generative models, with a particular focus on diffusion models and their theoretical foundations.
-
Neural Network Subspace Dynamics in Continual Learning
Investigated whether catastrophic forgetting occurs primarily in the bulk subspace of the Hessian matrix, designing optimization algorithms to isolate learning in different subspaces. -
Extrapolation and Distributional Robustness for Climate Downscaling (No public repository)
Researched transferability of statistical downscaling methods to unseen Global Climate Models, implementing (variance-reducing) generative models (elucidated diffusion model and engression MLPs) for high-resolution climate predictions. -
Machine Learning for Healthcare:
Built interpretable models for medical diagnosis and employed various machine learning architectures for arrhythmia classification.
I'm always open to collaborating on interesting problems at the intersection of machine learning theory and real-world applications. Feel free to reach out!