English

Inference Without Compatibility

Statistics Theory 2020-01-23 v2 Statistics Theory

Abstract

We consider hypotheses testing problems for three parameters in high-dimensional linear models with minimal sparsity assumptions of their type but without any compatibility conditions. Under this framework, we construct the first n\sqrt{n}-consistent estimators for low-dimensional coefficients, the signal strength, and the noise level. We support our results using numerical simulations and provide comparisons with other estimators.

Keywords

Cite

@article{arxiv.1903.06295,
  title  = {Inference Without Compatibility},
  author = {Michael Law and Ya'acov Ritov},
  journal= {arXiv preprint arXiv:1903.06295},
  year   = {2020}
}
R2 v1 2026-06-23T08:08:47.096Z