Abstract
In this paper, we present a simple, yet useful, concentration result concerning random (weighted) projections in high dimensional spaces. As one application, we prove a general concentration result for random quadratic forms, which extended a classical result of Hanson and Wright and improved several recent results. In another application, we show that the infinity norm of most unit eigenvectors of a random $\pm 1$ matrix is of optimal order $O(\sqrt {\log n/n})$, answering a question of Dekel, Lee and Linial and sharpening various earlier estimates. In fact, the estimate holds for a large class of random matrices. As a by-product, we obtain an estimate on the threshold for the local semi-circle law. This estimate is tight up to a $\sqrt {\log n}$ factor. It is an intriguing problem to determine if this factor is necessary.

This publication has 20 references indexed in Scilit: