A MAXIMUM-ENTROPY APPROACH TO ESTIMATION AND INFERENCE IN DYNAMIC-MODELS OR COUNTING FISH IN THE SEA USING MAXIMUM-ENTROPY

In this paper we consider estimation problems based on dynamic discret e time models. The first problem involves noisy state observations, wh ere the state equation and the observation equation are nonlinear. The objective is to estimate the unknown parameters of the state and obse rvation equations and the unknown values of the state variable. Next w e consider the problem of estimating the parameters of the objective f unction and of the state equation in a linear-quadratic control proble m. In each case, given time series observations, we suggest a nonlinea r inversion procedure that permits the unknown underlying parameters t o be estimated. Examples are presented to suggest the operational natu re of the results.