AĞIRLIKLI DOĞRUSAL OLMAYAN REGRESYON YÖNTEMİ VE BUNUN OPTİMİZASYON PROBLEMLERİNE UYGULANIŞI

İn this article, weighted nonhnear regression method (the variance of a l i distur bance of the variables i n a regression equation is not constant) and i t s a p p l i c a t i o n to the o p t i m i z a t i o n problems is presented. This approach is an i t e - r a t i v e l i n e a r i z a t i o n method i n which the nonlinear equations (constraints) are linear i z e d (using a Taylor series expansion) about the i n i t i al vectors of coefficients (parametersî and variables 1dependet and independentî. Then, weighted least squares is performed on this linear equations, generating a ne w vectors, The nonlinear equations are linearized about these vectors, weighted least squures is again performed t o generate new vectors, This i t e r a t i v e process is repeated i m i i l the vectors do not change substantially after each new weighted least squares regression. The f i n a l lineariz a t i o n of the i t e r a t i v e process w i l l provide some i n f o r m a t i - on. Such as; the values of coefficients and variables, t h e i r standart errors, and t h a t the nonlinear regression fıts the data. On the other hand a computer programı is also pre pared for t h i s purpose.