Universality of Benign Overfitting in Binary Linear Classification

Piotr Zwiernik

Serra Húnter Professor

I work on graphical models, covariance estimation, high-dimensional and tensor-valued data, as well as geometric methods in modern statistics and deep learning.