Weight coefficients in the weighted least squares method

Статья в журнале
Bychkov I. V., Zorkaltsev V.I., Kazazaeva A. V.
Numerical Analysis and Applications
Numerical Analysis and Applications. Vol.8. No.3. P.223-224.
2015
In this paper, a problem of estimating some parameters of linear mathematical models is considered. It is found that by choosing weight coefficients in the least squares method, one can obtain solutions by minimizing any penalty functions of a wide class, including Holder norms. It is proved that a set of solutions obtained by varying the weight coefficients in the least squares method is bounded. Some practical applications of the theoretical results are illustrated by an ecological-mathematical model as an example.

Библиографическая ссылка

Bychkov I. V., Zorkaltsev V.I., Kazazaeva A. V.  Weight coefficients in the weighted least squares method // Numerical Analysis and Applications. Vol.8. No.3. 2015. P.223-224. DOI: 10.1134/S1995423915030039 http://link.springer.com/article/10.1134%2FS1995423915030039
SCOPUS
x
x