2 problems

0 votes0 replies0 views
Local learnability conjecture for over-parameterized model classes
The paper considers pNML universal learning and the notion of learnability for model classes, with linear regression as the explicitly solved example. Local learnability conjecture…
Conjectureopenlinear-regression over-parameterization universal-learningarXiv posted 2 days ago
0 votes0 replies0 views
Data-derived convergence as local simplicity for rich model classes
A data-derived convergence property is a convergence notion in which the data determine when a model class is being approximated well, lying between uniform and pointwise convergen…
Conjectureopeninformation-theory universal-learningarXiv posted 2 days ago

Local learnability conjecture for over-parameterized model classes