Сравнение между нелинейния метод на най-малките квадрати и линеаризацияta на дробно-линеен регресионен модел
Comparison Between Nonlinear Least Square Method and Linearizations of Saturation Growth Model

Author(s): Deyan Mihaylov
Subject(s): Economy, ICT Information and Communications Technologies
Published by: Икономически университет - Варна
Keywords: Approximation; Least Square Method; Linearization; Saturation-Growth Model
Summary/Abstract: The determination of the regression parameters is one of the tasks of regression analysis. They are easily obtained in the case of linear dependence. The nonlinear models require applying more complicated methods. To avoid this it is often recommended to use linearization of the regression equation. This paper shows that this approach leads to increase in the bias of regression parameters. The example of saturation-growth model of Michaelis-Menten equation is used. The parameters of the linearization plots of Lineweaver-Burk, Eadie-Hofstee and Hanes-Wolf are biased from the parameters which are determined by nonlinear least square method.

• Details
• Contents