Abstract:Shuffled regression and unlinked regression represent intriguing challenges that have garnered considerable attention in many fields, including but not limited to ecological regression, multi-target tracking problems, image denoising, etc. However, a notable gap exists in the existing literature, particularly in vanishing noise, i.e., how the rate of estimation of the underlying signal scales with the error variance. This paper aims to bridge this gap by delving into the monotone function estimation problem under vanishing noise variance, i.e., we allow the error variance to go to $0$ as the number of observations increases. Our investigation reveals that, asymptotically, the shuffled regression problem exhibits a comparatively simpler nature than the unlinked regression; if the error variance is smaller than a threshold, then the minimax risk of the shuffled regression is smaller than that of the unlinked regression. On the other hand, the minimax estimation error is of the same order in the two problems if the noise level is larger than that threshold. Our analysis is quite general in that we do not assume any smoothness of the underlying monotone link function. Because these problems are related to deconvolution, we also provide bounds for deconvolution in a similar context. Through this exploration, we contribute to understanding the intricate relationships between these statistical problems and shed light on their behaviors when subjected to the nuanced constraint of vanishing noise.

Convergence Rates of Multivariate Regression Estimators with Errors-In-Variables

Pathwise Derivatives for Multivariate Distributions

Asymptotic Properties of the Least Squares Estimator in Local to Unity Processes with Fractional Gaussian Noise

Convergence Rate of Wavelet Estimator in Semiparametric Models with Dependent Mh(∞)Error Process

Estimation of Convergence Rate for Multi-Regression Learning Algorithm

Minimax Optimal rates of convergence in the shuffled regression, unlinked regression, and deconvolution under vanishing noise

Convergence rates of empirical Bayesian estimation in a class of linear models

Precise Error Analysis of Regularized M-estimators in High-dimensions

Convergence rates of non-stationary and deep Gaussian process regression

Convergence rates in the Prokhorov metric for assessing uncertainty in ill-posed problems

Convergence rates for estimating multivariate scale mixtures of uniform densities

Minimax-rate Adaptive Nonparametric Regression with Unknown Correlations of Errors

Achieving High Convergence Rates by Quasi-Monte Carlo and Importance Sampling for Unbounded Integrands

On Wavelet Regression with Long Memory Infinite Moving Average Errors

Wavelet-based estimation of regression function with strong mixing errors under fixed design

Convergence Rates for the Maximum A Posteriori Estimator in PDE-Regression Models with Random Design

Convergence Analysis of MCMC Algorithms for Bayesian Multivariate Linear Regression with Non‐Gaussian Errors

Errors-in-variables regression for mixed Euclidean and non-Euclidean predictors

Minimax rates of convergence for nonparametric location-scale models

Rates of convergence in conditional covariance matrix with nonparametric entries estimation

Local Polynomial Estimator of the Regression Function in Autoregressive Models with Errors in Variables