hrcak mascot   Srce   HID

Izvorni znanstveni članak

A New Median - Based Robust Regression Method

Costel Sárbu ; Department of Analytical Chemistry, »Babes - Bolyai« University, RO - 3400 Cluj - Napoca, Roumania

Puni tekst: engleski, pdf (8 MB) str. 795-801 preuzimanja: 182* citiraj
APA 6th Edition
Sárbu, C. (1997). A New Median - Based Robust Regression Method. Croatica Chemica Acta, 70 (3), 795-801. Preuzeto s https://hrcak.srce.hr/135647
MLA 8th Edition
Sárbu, Costel. "A New Median - Based Robust Regression Method." Croatica Chemica Acta, vol. 70, br. 3, 1997, str. 795-801. https://hrcak.srce.hr/135647. Citirano 04.03.2021.
Chicago 17th Edition
Sárbu, Costel. "A New Median - Based Robust Regression Method." Croatica Chemica Acta 70, br. 3 (1997): 795-801. https://hrcak.srce.hr/135647
Harvard
Sárbu, C. (1997). 'A New Median - Based Robust Regression Method', Croatica Chemica Acta, 70(3), str. 795-801. Preuzeto s: https://hrcak.srce.hr/135647 (Datum pristupa: 04.03.2021.)
Vancouver
Sárbu C. A New Median - Based Robust Regression Method. Croatica Chemica Acta [Internet]. 1997 [pristupljeno 04.03.2021.];70(3):795-801. Dostupno na: https://hrcak.srce.hr/135647
IEEE
C. Sárbu, "A New Median - Based Robust Regression Method", Croatica Chemica Acta, vol.70, br. 3, str. 795-801, 1997. [Online]. Dostupno na: https://hrcak.srce.hr/135647. [Citirano: 04.03.2021.]

Sažetak
Most instrumental determinations require computation of linear calibration parameters based on a small number of chemical standards. The median - based regression methods are robust methods, which means that they are not sensitive to outliers or other violations of the assumptions of the usual normal model. This contrasts with the conventional regression method, which minimizes the sum of residuals squares. It is demonstrated that the performance of the proposed method, namely mean median exceeds that of the ordinary least squares method, and often equals that of well known robust median methods.

Hrčak ID: 135647

URI
https://hrcak.srce.hr/135647

Posjeta: 273 *