English

Finite-sample properties of the trimmed mean

Statistics Theory 2025-01-08 v1 Probability Statistics Theory

Abstract

The trimmed mean of nn scalar random variables from a distribution PP is the variant of the standard sample mean where the kk smallest and kk largest values in the sample are discarded for some parameter kk. In this paper, we look at the finite-sample properties of the trimmed mean as an estimator for the mean of PP. Assuming finite variance, we prove that the trimmed mean is ``sub-Gaussian'' in the sense of achieving Gaussian-type concentration around the mean. Under slightly stronger assumptions, we show the left and right tails of the trimmed mean satisfy a strong ratio-type approximation by the corresponding Gaussian tail, even for very small probabilities of the order ence^{-n^c} for some c>0c>0. In the more challenging setting of weaker moment assumptions and adversarial sample contamination, we prove that the trimmed mean is minimax-optimal up to constants.

Keywords

Cite

@article{arxiv.2501.03694,
  title  = {Finite-sample properties of the trimmed mean},
  author = {Roberto I. Oliveira and Paulo Orenstein and Zoraida F. Rico},
  journal= {arXiv preprint arXiv:2501.03694},
  year   = {2025}
}

Comments

37 pages