Biometrical Letters vol. 48(2), 2011, pp. 101-111
|
EMPIRICAL COMPARISON OF ITERATIVE METHODS IN LOGISTIC MODELS BASED ON THREE REAL EXPERIMENTS AND SIMULATED DATA Ewa Bakinowska, Bogna Zawieja Department of Mathematical and Statistical Methods, Poznan University of Life Sciences, Wojska Polskiego 28, 60-637 Poznań, Poland; e-mail: ewabak@up.poznan.pl |
The estimation of unknown parameters in a logistic model with multinomial distribution is usually carried out using iterative methods: the Newton-Raphson or Fisher scoring method. These methods often differ in the number of iterations, which is connected with the type of model used. In this paper these two iterative methods are compared empirically. Data from three real experiments were used for the analysis. Each experiment was described by a different logistic model. Additionally, calculations were carried out for two experiments expanded with simulated data.
ear emergence, grain contamination, Fisher scoring method, hatchability of chicks, logistic model, maximum likelihood method, Newton-Raphson method
