Extends high-dimensional KRR to product kernels, proving convergence rates that recover minimax optimality for source condition s ≤ 1, saturation for s > 1, and multiple-descent phenomena with respect to sample size n.
Deep equals shallow for ReLU networks in kernel regimes
3 Pith papers cite this work, alongside 9 external citations. Polarity classification is still indexing.
3
Pith papers citing it
9
external citations · external index
verdicts
UNVERDICTED 3representative citing papers
Temporal correlations from lazy random walks enable efficient SGD learning of k-juntas via temporal-difference loss on ReLU networks, achieving linear sample complexity in d.
A method is given to determine eigenvalue decay rates of NTK and related kernels on general domains, leading to minimax optimality results for wide neural networks under smoothness assumptions on the target function.
citing papers explorer
-
Large Dimensional Kernel Ridge Regression: Extending to Product Kernels
Extends high-dimensional KRR to product kernels, proving convergence rates that recover minimax optimality for source condition s ≤ 1, saturation for s > 1, and multiple-descent phenomena with respect to sample size n.
-
The Benefits of Temporal Correlations: SGD Learns k-Juntas from Random Walks Efficiently
Temporal correlations from lazy random walks enable efficient SGD learning of k-juntas via temporal-difference loss on ReLU networks, achieving linear sample complexity in d.
-
On the Eigenvalue Decay Rates of a Class of Neural-Network Related Kernel Functions Defined on General Domains
A method is given to determine eigenvalue decay rates of NTK and related kernels on general domains, leading to minimax optimality results for wide neural networks under smoothness assumptions on the target function.