PhD student, University of Toronto
1 paper at NeurIPS 2025
We prove the asymptotic normality of PCA on the Grassmannian, and derive a tight non-asymptotic bound on its excess risk using self-concordance.
