3 problems

0 votes0 replies0 views
Conjecture that long-horizon upper bounds are not tight for smooth realizable SCO
Let $ceta$ be the step size, $T$ the number of iterations, and $n$ the sample size in smooth realizable stochastic convex optimization. Consider the regime $ceta T = cOmega(n)$ , co…
Conjectureopengeneralization-bounds gradient-descent stochastic-convex-optimizationarXiv posted 2 days ago
0 votes0 replies1 view
Xu's adaptive-partitioning conjecture for robust generalization bounds
Let $K$ denote the size of the partition (or covering number) used in an algorithmic robustness bound, and let $\boldsymbol{adaptive\ partitioning}$ mean choosing that partition in…
Conjectureopenalgorithmic-robustness generalization-bounds learning-theoryarXiv posted 2 days ago
0 votes0 replies1 view
The strengthened Fourier-class inclusion conjecture for PQC-based models
Strengthened Fourier-class inclusion conjecture. For any PQC-based model, one has
Conjectureopenfourier-analysis generalization-bounds quantum-machine-learningarXiv posted 2 days ago

Conjecture that long-horizon upper bounds are not tight for smooth realizable SCO