Lineární regresní model a vážená metoda nejmenších čtverců

Warning

This publication doesn't include Faculty of Sports Studies. It includes Faculty of Science. Official publication website can be found on muni.cz.

Title in English	Linear Regresion Model and Weighted Least
Authors	MICHÁLEK Jaroslav
Year of publication	2000
Type	Article in Proceedings
Conference	XIV. letná škola biometriky, BIOMETRICKÉ METÓDY A MODELY V PÔDOHOSPODÁRSKEJ VEDE, VÝSKUME A VÝUKE
MU Faculty or unit	Faculty of Science
Citation
Field	Applied statistics, operation research
Keywords	Linear regression model; Method of least squares; BLUE; General linear hypothesis; DFITS; Robust regression weights.
Description	Linear regression model and properties of the estimates of the model parameters are described in the first part of the paper. Then the test of the general linear hypothesis is involved. The second part is devoted to the generalized and weighted least squares method and to robust procedures in regression analysis. The robust weights which depend on combined effect of leverage and Studentized residual and enable to eliminate outliers and high leverage points in given biometrical data are described in the paper.
Related projects:	Functional-differential equations and mathematical-statistical models