Department of Mathematical Sciences
|Thursday, November 19, 2015
|1:15pm to 2:15pm
|Vladislav Kargin, Binghamton University
|On singular Values of the Reduced-Rank Regression
I will talk about the multivariate linear regression in the case when the number of responses and predictors is large and comparable with the number of observations. One approach uses the reduced-rank regression, in which the rank of the coe fficient matrix is assumed to be small. I will relate this problem to certain problems in random matrix theory. I will also talk about algorithms for the model rank selection and about consistent estimation of the singular values of the coefficient matrix.