Adaptive polynomial least squares slope estimates of noisy data. Issue 7 (3rd October 2021)

• Record Type: Journal Article
• Authors: Van Hecke, Tanja
• Is Part Of: Journal of statistics & management systems. Volume 24:Issue 7(2021)
• Page Start: 1543
• Page End: 1549
• Publication Date: 2021-10-03
• Subjects: 65D10 -- 62J02 -- 94A12; Derivative approximation -- Least squares differences -- SNR; Statistics -- Periodicals; Mathematical models -- Periodicals; Mathematical models; Statistics; Periodicals; 519.5
• DOI: 10.1080/09720510.2021.1914427
• Languages: English
• ISSNs: 0972-0510

Abstract

Abstract: Noise creates a harmful transformation on the estimation of the derivative of a signal. In this paper, we suggest an easy to implement and adequate approximation method for the derivative function of a signal with noise. It is based on polynomial regression constructed with least squares differences, but with a degree depending on the signal-to-noise ratio. Numerical results suggest similar accuracy compared to established methods as the Savitzsky-Golay smoothing method and the smoothing method based on wavelet transforms.