Enhanced Mean Estimation Using Memory Type Estimators with Dual Auxiliary Variables

Accepted - January 2025

Abstract

The study investigates how Exponentially Weighted Moving Average (EWMA) is mathematically expressed, what parameter values work best for mean estimation with different data types, and how well it detects slight changes in the process mean. This study suggests a family of memory-type estimators that use two auxiliary variables to estimate the population mean. The suggested class of estimators’ mean squared error has been calculated up to the first degree of approximation. Compared to estimators in the literature that use two auxiliary variables, the suggested memory type estimators are found to be more effective. To support the theoretical findings, an empirical and simulation investigation are conducted.