The ℓ¹-regularized empirical risk minimizer achieves minimax-optimal convergence rate n^{-r/(1+b-br)} for nonlinear statistical inverse learning under variational source conditions and polynomial effective-dimension decay.
Vector valued reproducing kernel Hilbert spaces and universality.Analysis and Applications, 8(01):19–61
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
stat.ML 2representative citing papers
Establishes near-optimal dimension-independent convergence rates for regularized SGD with operator-valued kernels in statistical inverse problems for operator learning.
citing papers explorer
-
Statistical inverse learning and $\ell^1$-regularization
The ℓ¹-regularized empirical risk minimizer achieves minimax-optimal convergence rate n^{-r/(1+b-br)} for nonlinear statistical inverse learning under variational source conditions and polynomial effective-dimension decay.
-
Learning Operators by Regularized Stochastic Gradient Descent with Operator-valued Kernels
Establishes near-optimal dimension-independent convergence rates for regularized SGD with operator-valued kernels in statistical inverse problems for operator learning.