| Sumario: | In this paper, the author generalizes the modified Newton method previously applied to the computation of full-information maximum likelihood estimates of parameters of a system of linear structural equations to the case of a system of nonlinear structural equations. The success of that method for linear systems has stimulated author's present attempt to generalize it for nonlinear systems. The subject of maximum likelihood estimation of nonlinear simultaneous equation systems has been studied by authors. The estimation equations for nonlinear systems are derived, under the assumptions that each structural equation contains a distinct set of parameters, that the parameters are not subject to any linear restrictions, and that the (additive) residuals are serially uncorrelated. In order to appreciate the nature and the difficulty of estimating the parameters of non-linear equations, it is useful to present the estimating equations when any structural equation is linear. It is also of practical importance to do so, since linear structural equations are often encountered in practice, and one would wish to exploit the linearity to simplify computations.
|
|---|