1 problem

0 votes0 replies1 view
Conjecture on tightening the dimension dependence of the Hessian approximation bound
Let $f”_1(w_k)$ and $f”(w_k)$ be the empirical and population Hessians appearing in the preceding high-probability bound, and let $d$ denote the dimension. The bound … contains a d…
Conjectureopenconcentration-inequalities distributed-optimization self-concordant-lossesarXiv posted 3 days ago

Conjecture on tightening the dimension dependence of the Hessian approximation bound