Adaptive Noisy Clustering

TL;DR

This paper introduces an adaptive noisy clustering method that employs a deconvolution-based $k$-means approach with a data-driven bandwidth selection rule, achieving fast convergence rates despite noise.

Contribution

It proposes a novel adaptive noisy $k$-means clustering method with a new bandwidth selection rule based on empirical risk comparison, applicable to various $M$-estimation problems.

Findings

Achieves fast convergence rates for excess risk.

Develops a data-driven bandwidth selection rule (ERC).

Demonstrates applicability to multiple statistical problems.

Abstract

The problem of adaptive noisy clustering is investigated. Given a set of noisy observations $Z_i=X_i+\epsilon_i$, $i=1,...,n$, the goal is to design clusters associated with the law of $X_i$'s, with unknown density $f$ with respect to the Lebesgue measure. Since we observe a corrupted sample, a direct approach as the popular {\it $k$-means} is not suitable in this case. In this paper, we propose a noisy $k$-means minimization, which is based on the $k$-means loss function and a deconvolution estimator of the density $f$. In particular, this approach suffers from the dependence on a bandwidth involved in the deconvolution kernel. Fast rates of convergence for the excess risk are proposed for a particular choice of the bandwidth, which depends on the smoothness of the density $f$. Then, we turn out into the main issue of the paper: the data-driven choice of the bandwidth. We state an adaptive upper bound for a new selection rule, called ERC (Empirical Risk Comparison). This selection rule is based on the Lepski's principle, where empirical risks associated with different bandwidths are compared. Finally, we illustrate that this adaptive rule can be used in many statistical problems of $M$-estimation where the empirical risk depends on a nuisance parameter.