Pointwise minimax lower bounds under a multi-index constraint: the price of an unknown orientation

Abstract

We consider nonparametric regression with random design and non-Gaussian noise, where the regression function has the multi-index form F(x)=f(Θx) with an unknown link f:RR^m→RR in an anisotropic Hölder class and an unknown matrix Θ with m≤ d orthonormal rows. For estimation of F at a fixed point we prove that the minimax risk is bounded from below by a multiple of L^1/(2γ+1)(n⁻¹ln n)^γ/(2γ+1), where γ is the harmonic mean of the smoothness indices of the link divided by m. For m<d this extends a Gaussian white noise result to regression. The main new case is m=d, where no dimension reduction takes place: the logarithmic factor still appears as soon as the link is anisotropic and the orientation of its axes is unknown. In contrast, when the link is isotropic with smoothness index at most one, the class is rotation invariant up to a constant and the classical rate without logarithm is attained. The logarithmic loss is therefore caused by the unknown orientation of the directions in which the function is least smooth

Quadratic Functional Estimation and Pointwise Inference from Paired Measurements with Unknown Errors Under Polynomial Lower Bounds
Lulah Alnaji
Axioms27 Oct 2026Peer-reviewedCC BY 4.0StatisticsCrossref
Minimax rates for pointwise CATE estimation: the role of design smoothness
Yuxuan Hou, Zhenghan Li
arXiv13 Oct 2026PreprintarXiv licenceFull textStatisticsarXiv
Minimax Additive Regression under Unknown Dependent Designs
Baptiste Ferrere, Fabrice Gamboa, Jean-Michel Loubes
arXiv3 Oct 2026PreprintCC BY 4.0Full textStatisticsarXiv
On some bounds for cardinal invariants of the space of quasicontinuous functions with the topology of pointwise convergence
Mandeep Kumar, Brij Kishore Tyagi
arXiv24 Oct 2026PreprintCC BY 4.0Full textGeometry & topologyarXiv
High-Probability Minimax Lower Bounds
Tianyi Ma, Kabir Aladin Verchand, Richard J. Samworth
Statistical Science31 Jul 2026Peer-reviewedLicence not statedFull text24 citedStatisticsOpenAlex

Cite

BibTeX RIS

BibTeX

@misc{serdyukova2026pointwise,
  title = {{Pointwise minimax lower bounds under a multi-index constraint: the price of an unknown orientation}},
  author = {Nora Serdyukova},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.19208},
  eprint = {2610.19208},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.19208},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - Pointwise minimax lower bounds under a multi-index constraint: the price of an unknown orientation
AU  - Nora Serdyukova
PY  - 2026
DA  - 2026/10/13
PB  - arXiv
DO  - 10.48550/arxiv.2610.19208
UR  - https://arxiv.org/abs/2610.19208
AB  - We consider nonparametric regression with random design and non-Gaussian noise, where the regression function has the multi-index form F(x)=f(Θx) with an unknown link f:RR^m→RR in an anisotropic Hölder class and an unknown matrix Θ with m≤ d orthonormal rows. For estimation of F at a fixed point we prove that the minimax risk is bounded from below by a multiple of L^1/(2γ+1)(n^−1ln n)^γ/(2γ+1), where γ is the harmonic mean of the smoothness indices of the link divided by m. For m<d this extends a Gaussian white noise result to regression. The main new case is m=d, where no dimension reduction takes place: the logarithmic factor still appears as soon as the link is anisotropic and the orientation of its axes is unknown. In contrast, when the link is isotropic with smoothness index at most one, the class is rotation invariant up to a constant and the classical rate without logarithm is attained. The logarithmic loss is therefore caused by the unknown orientation of the directions in which the function is least smooth
ER  -

Abstract

Related

Cite