Continuous Optimization Seminar - Sina Baghal

Title: High dimensional probability: Estimation in high dimensions

Abstract:

In this talk, we discuss high dimensional estimation problems. The aim is to estimate some point x in a given ambient bounded space K using a small number of independent random observations. We first discuss the notion of Gaussian width for bounded sets and give some bounds for its value on some examples. Then, using Gaussian width, we discuss estimation from noisy linear observations; namely when x lies in the intersection of K and the epsilon vicinity of some random affine manifold. If time permits, we also mention similar results on sparse recovery.