My research lies at the intersection of statistics, random matrix theory, and free probability. I am primarily interested in studying high-dimensional neural network models from the perspective of random matrix theory. Specifically, I investigate the relationships between the eigenvalue density of random matrices and the performance of neural networks. This viewpoint allows for the consideration of various neural network variants, including the case of data with a variance profile.

Research Interests

Deep Learning / Random Features Model / Linear regression / High Dimension /Ridge regression / Data with variance profile (Heteroscedastic) / Random Matrix Theory / Free Probability / Traffic Theory / Graph Theory / Combinatorics