Michael J. Feldman

Logo

I am a postdoctoral researcher in the Department of Statistics at Stanford University. I completed my Ph.D. here at Stanford in June 2023, advised by David Donoho. My undergraduate degree is in Applied and Computational Mathematics from Caltech.

My research interests include high-dimensional statistics, random matrix theory, covariance estimation, and tensor methods.

Contact

email: “lastname”6@stanford.edu
office: 244 Sequoia Hall

Publications and preprints

Spectral Properties of Elementwise-Transformed Spiked Matrices
M. J. Feldman, arXiv preprint, 2023.
arXiv

Sharp Recovery Thresholds of Tensor PCA Spectral Algorithms
D. L. Donoho and M. J. Feldman, to appear in NeurIPS 2023.
pdf

Optimal Eigenvalue Shrinkage in the Semicircle Limit
D. L. Donoho and M. J. Feldman, arXiv preprint, 2022.
arxiv

Spiked Singular Values and Vectors under Extreme Aspect Ratios
M. J. Feldman, Journal of Multivariate Analysis 196:105187, 2023.
jmva arXiv