Random matrices: The distribution of the smallest singular values
Abstract
Let be a real-valued random variable of mean zero and variance 1. Let denote the random matrix whose entries are iid copies of and denote the least singular value of . ( is also usually interpreted as the least eigenvalue of the Wishart matrix .) We show that (under a finite moment assumption) the probability distribution is {\it universal} in the sense that it does not depend on the distribution of . In particular, it converges to the same limiting distribution as in the special case when is real gaussian. (The limiting distribution was computed explicitly in this case by Edelman.) We also proved a similar result for complex-valued random variables of mean zero, with real and imaginary parts having variance 1/2 and covariance zero. Similar results are also obtained for the joint distribution of the bottom singular values of for any fixed (or even for growing as a small power of ) and for rectangular matrices. Our approach is motivated by the general idea of "property testing" from combinatorics and theoretical computer science. This seems to be a new approach in the study of spectra of random matrices and combines tools from various areas of mathematics.
Cite
@article{arxiv.0903.0614,
title = {Random matrices: The distribution of the smallest singular values},
author = {Terence Tao and Van Vu},
journal= {arXiv preprint arXiv:0903.0614},
year = {2009}
}