A Vector Bernstein Inequality for Self-Normalized Martingales

Ingvar Ziemann
2024-12-30
Abstract:We prove a Bernstein inequality for vector-valued self-normalized martingales. We first give an alternative perspective of the corresponding sub-Gaussian bound due to \cite{abbasi2011improved} via a PAC-Bayesian argument with Gaussian priors. By instantiating this argument to priors drawn uniformly over well-chosen ellipsoids, we obtain a Bernstein bound.
Probability
What problem does this paper attempt to address?