Chunked-and-Averaged Estimators for Vector Parameters

TL;DR

This paper introduces and analyzes chunked-and-averaged estimators for vector parameters, extending previous univariate and IID sampling studies to non-IID contexts, providing a divide-and-conquer approach for complex data.

Contribution

It extends CA estimators to vector parameters and non-IID sampling, broadening their applicability in statistical estimation.

Findings

Effective for vector parameters under non-IID sampling

Generalizes previous IID univariate results

Provides theoretical foundations for CA estimators in complex settings

Abstract

A divide-and-conquer method for parameter estimation is the chunked-and-averaged (CA) estimator. CA estimators have been studied for univariate parameters under independent and identically distributed (IID) sampling. We study the CA estimators of vector parameters and under non-IID sampling.