3 problems

0 votes0 replies0 views
The dimensional scaling conjecture for random feature method convergence
Dimensional scaling conjecture. In higher dimensions, the convergence rate can be roughly estimated by replacing $N$ with $N^{1/d}$ in the one-dimensional convergence rate.
Conjectureopennumerical-analysis random-feature-methods spectral-convergencearXiv posted 2 days ago
0 votes0 replies0 views
Sharp $\delta^{1/2}$ convergence rate for thin-domain Reissner\Mindlin systems
Let $(A_\delta)_\delta$ be the family of non-negative self-adjoint operators associated with the ReissnerMindlin system on thin domains $\Omega_\delta$ , and let…
Conjectureopenreissner-mindlin-systems spectral-convergence thin-domainsarXiv posted 2 days ago
0 votes0 replies0 views
Conjecture on data-driven stabilization and spectral convergence for combinatorial Hodge-Laplacians
Let $(X,d,cmu)$ be a metric-measure space, and let the sample size increase for the data-driven procedures estimating the real cohomology groups of $(X,d,cmu)$ . Consider the Hodge-…
Conjectureopenmetric-measure-spaces spectral-convergence topological-data-analysisarXiv posted 2 days ago

The dimensional scaling conjecture for random feature method convergence