Estimating Non-Linear Regression Parameters Using Denoised Variables

Abstract

The observed data from various fields are frequently characterized by measurement error and this has been challenging problem to construct consistent estimators of the parameters in a nonlinear regression model.This study uses simulated data under three (3) sample sizes (i.e 32,256 and 1024) applying Kernel, Wavelet and Polynomial Spline  on noisy data in two approaches (i.e denoising only the explanatory variables and denoising both dependent and explanatory variables). The study reveals the performance of denoised nonlinear estimators under different sample sizes for each denoising approach and comparison was made using the mean squared error criterion. The result of the studies shows that the denoised nonlinear least squares estimator (DNLS) is the best under each sample size considered.

Keywords: Denoising, Measurement Error, Monte-Carlo Simulation, Non-linear regression, Production model, Smoothers


Article Review Status: Published

Pages: 61-71 (Download PDF)

Creative Commons Licence
This work by European American Journals is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License