A vector-contraction inequality for Rademacher complexities using $p$-stable variables
This is an incremental theoretical extension for researchers in statistical learning theory.
The paper tackles the problem of extending a vector-contraction inequality for Rademacher complexities by replacing sub-gaussian variables with p-stable variables for 1<p<2, resulting in a theoretical generalization of prior work.
Andreas Maurer in the paper "A vector-contraction inequality for Rademacher complexities" extended the contraction inequality for Rademacher averages to Lipschitz functions with vector-valued domains; He did it replacing the Rademacher variables in the bounding expression by arbitrary idd symmetric and sub-gaussian variables. We will see how to extend this work when we replace sub-gaussian variables by $p$-stable variables for $1<p<2$.