Economists currently deal with the modeling of nonlinear processes, having complex structure and large dimensionality. The traditional modeling methods, such as nonlinear regression analysis, generally fail in providing appropriate solutions to this problem, especially in the cases when the functional form of the model is unknown and has to be guessed.
This paper is intended to give a summary description of an alternative technique, the fuzzy logic modeling method, and to promote it as a powerful tool for modeling producer behavior. In order to provide a comparative analysis, we test for estimation accuracy both in the fuzzy logic model and in the parametric model based on translog function that is accepted as one of the most flexible functional form. The two methods are applied to the same data set (reported by Christensen and Greene, 1976). Finally, we describe a MATLAB implementation of a flexible technique for fuzzy logic model identification and system optimization.