HDI Lab seminar 'Sharp Deviation Bounds for Quadratic Forms and Their Use in Machine Learning'

On June 1 a seminar of the International Laboratory of Stochastic Algorithms and High-Dimensional Inference will be held at HSE University. Vladimir Spokoiny (HSE University, RAS Institute for Information Transmission Problems, Humboldt University, WIAS Berlin) will speak on 'Sharp deviation bounds for quadratic forms and their use in machine learning'.

Abstract:

Under a mild assumption of linearity on the log-likelihood function, the tools of empirical processes are not necessary, the analysis of a general MLE can be reduced to a deviation bound for a quadratic form. This paper explains how the recent advances in Laplace approximation from [Spokoiny, 2022, 2023] can be used for obtaining sharp Gaussian-like finite sample bounds and for stating the prominent concentration phenomenon for the squared norm of a sub-gaussian vector. Some extensions and open problems will be discussed as well.

Start time: 16:20

Venue: 11 Pokrovsky Bulvar, room N508