engleski

Estimation of optimal parameters of generalized logistic model-function in the discrete $L_p$-norms

We consider the existence problem for optimal parameters of the generalized logistic function approximating the given data $(w_i,t_i,f_i)$, $i=1,ldots,m,$ in the sense of the least $L_p$-norms, $1leq p<infty$. Under the natural condition on the data, we prove the existence of such optimal parameters. Also, we give a few examples in which we estimate the optimal parameters of the generalized logistic model-function in the sense of the discrete least $L_p$-norm (for some values $p$), by means of the iteratively reweighted least squares method and the linearization of the generalized logistic function.

generalized logistic function; $L_p$-norm; parameter estimation; iteratively reweighted least squares method

nije evidentirano