A Fast Spectral Algorithm for Mean Estimation with Sub-Gaussian Rates

TL;DR

This paper introduces a spectral algorithm for mean estimation of heavy-tailed distributions that achieves optimal sub-Gaussian error bounds with significantly improved computational efficiency over previous SDP-based methods.

Contribution

The authors develop a novel spectral algorithm that surpasses SDP-based approaches in runtime, leveraging eigenvector computations for efficient mean estimation under heavy-tailed data.

Findings

Achieves optimal sub-Gaussian error bounds.

Runs in time rac{rac{n^2 d}{ ilde{}}}

Improves previous runtime from rac{rac{n^{3.5}+ n^2d}{ ilde{}}}.

Abstract

We study the algorithmic problem of estimating the mean of heavy-tailed random vector in $\mathbb{R}^d$, given $n$ i.i.d. samples. The goal is to design an efficient estimator that attains the optimal sub-gaussian error bound, only assuming that the random vector has bounded mean and covariance. Polynomial-time solutions to this problem are known but have high runtime due to their use of semi-definite programming (SDP). Conceptually, it remains open whether convex relaxation is truly necessary for this problem. In this work, we show that it is possible to go beyond SDP and achieve better computational efficiency. In particular, we provide a spectral algorithm that achieves the optimal statistical performance and runs in time $\widetilde O\left(n^2 d \right)$, improving upon the previous fastest runtime $\widetilde O\left(n^{3.5}+ n^2d\right)$ by Cherapanamjeri el al. (COLT '19). Our algorithm is spectral in that it only requires (approximate) eigenvector computations, which can be implemented very efficiently by, for example, power iteration or the Lanczos method. At the core of our algorithm is a novel connection between the furthest hyperplane problem introduced by Karnin et al. (COLT '12) and a structural lemma on heavy-tailed distributions by Lugosi and Mendelson (Ann. Stat. '19). This allows us to iteratively reduce the estimation error at a geometric rate using only the information derived from the top singular vector of the data matrix, leading to a significantly faster running time.