Maximum Entropy Estimation for Survey sampling

TL;DR

This paper introduces a maximum entropy approach to survey sampling calibration, framing it as an inverse problem and offering a new computational perspective and analysis of statistical properties.

Contribution

It extends traditional calibration methods by applying maximum entropy principles, providing a novel framework for estimating survey weights.

Findings

New maximum entropy calibration method proposed

Provides a computational framework for survey weight estimation

Analyzes statistical properties of the new method

Abstract

Calibration methods have been widely studied in survey sampling over the last decades. Viewing calibration as an inverse problem, we extend the calibration technique by using a maximum entropy method. Finding the optimal weights is achieved by considering random weights and looking for a discrete distribution which maximizes an entropy under the calibration constraint. This method points a new frame for the computation of such estimates and the investigation of its statistical properties.