4 problems

Improved packing conjecture. It may be possible to construct such a packing set with more careful control of the $\ell$-infinity norm, so that the resulting graphon estimation boun…

Let $A$ be the observed adjacency matrix, let $M$ denote the graphon-generated mean adjacency matrix, and let $\ell$ be the estimation loss. The computational optimality conjecture…

Let $W_0$ be a graphon, let $\rho_n>0$, and let $k$ be the rank parameter appearing in Proposition 1. The proposition gives convergence-rate bounds for graphon estimators over the…

Let $\rho_n$ denote the graph sparsity parameter, let $\rho_n^2/n^{\alpha}$ be the agnostic loss term arising for $alpha\leq 1$ when $r\asymp\rho_n$, and let $delta^2(\cdot,\cdot)$…