Naslov
On the $l_s$-norm generalization of the NLS method for the Bass model

Autori
Jukić, Dragan

Izvornik
European journal of pure and applied mathematics (1307-5543) 6 (2013), 4; 435-450

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
Bass model; diffusion; $l_s$-norm estimate; Least squares estimate; Existence problem; Data fitting

Sažetak
The best-known and widely used model in diffusion research is the Bass model. Estimation of its parameters has been approached in the literature by various methods, among which a very popular one is the nonlinear least squares (NLS) method proposed by Srinivasan and Mason. In this paper, we consider the $l_s$-norm $(1\leq s<\infty)$ generalization of the NLS method for the Bass model. Our focus is on the existence of the corresponding best $l_s$-norm estimate. We show that it is possible for the best $l_s$-norm estimate not to exist. As a main result, two theorems on the existence of the best $l_s$-norm estimate are obtained. One of them gives necessary and sufficient conditions which guarantee the existence of the best $l_s$-norm estimate.

Izvorni jezik
Engleski

Znanstvena područja
Matematika, Ekonomija

POVEZANOST RADA