Tuesday, June 12, 2018 - 16:15 in X-E0-218
Concentration of measure and functional inequalities
A talk in the SYMBol series by
Arthur Sinulis
Abstract: |
I will present the general concept of a concentration of measure, which was first used in the 70s in the theory of Banach spaces, and was developed in the 90s and 2000s by various authors. To this end, I will introduce some functional inequalities (initially proven in the context of Sobolev spaces) and hint at how to use these to obtain concentration results. I will end the talk by presenting an application: the Johnson-Lindenstrauss theorem, informally stating that almost any random linear mapping from a high to a low-dimensional space is an $\varepsilon$-isometry. |
Back