Worst-case L p-approximation of periodic functions using median lattice algorithms
math.NA, cs.NA
Submitted: 2026-03-05
Updated: 2026-10-29
License: http://creativecommons.org/licenses/by/4.0/
The gist: We study the worst-case approximation of multivariate periodic functions from the weighted Korobov space H d,α…
Terminology
Abstract
We study the worst-case approximation of multivariate periodic functions from the weighted Korobov space H d,α, γ with smoothness α>1/2 and weights γ in the Lebesgue norm L p([0,1] d) for 1 at most p at most infinity. We analyze a median lattice algorithm that reconstructs a truncated Fourier series by approximating the coefficients on a hyperbolic-cross-type index set using R rank-1 lattice sampling rules with independent randomly chosen generating vectors, and then aggregating the resulting coefficient estimators via the componentwise median. For an odd number of repetitions R and an odd prime lattice size N, where R depends logarithmically on N, we prove high-probability error bounds in both L infinity and L 2. Interpolation then yields the result for all p in [1, infinity]. In particular, with high probability, the worst-case L p-error satisfies, for any β>0, an upper bound of the following form C d,α,β, γ,τ,p,R 1 over 2+(12 - 1p)+ N- α+ (12 - 1p)+ + β, where (x)+ = x, 0, and the constant C d,α,β, γ,τ,p is independent of N and R. For p= infinity, C d,α,β, γ,τ, infinity is dimension-independent under the summability condition sum from j=1 to infinity γ j 1/(2α)< infinity. These results extend recent analyses of median-based lattice approximation in L 2 and complement related multiple-shift lattice approaches, showing that median aggregation yields nearly optimal L p-approximation rates (up to logarithmic factors and an arbitrarily small loss) in weighted Korobov spaces.
Sources
- Minimal Subsampled Rank-1 Lattices for Multivariate Approximation with Optimal Convergence Rate
- A lattice algorithm with multiple shifts for function approximation in Korobov spaces
- A note on approximation in weighted Korobov spaces via multiple rank-1 lattices
- Universal $L_2$-approximation using median lattice algorithms
- $L_2$-approximation using median lattice algorithms
Related papers
- Do physics-informed neural networks (PINNs) need to be deep? Shallow PINNs using the Levenberg-Marquardt algorithm
- A Neural-preconditioned Poisson Solver for Mixed Dirichlet and Neumann Boundary Conditions
- Advanced Linear Algebra with Applications - Part I (Numerical linear algebra for PDEs, machine learning, and data assimilation)
- Operator learning meets inverse problems: A probabilistic perspective
- Second-order consistency for learning chaotic dynamics via randomized Jacobian matching
- Windowed thinning and query complexity for the bouncy particle and Zigzag samplers