Approximation accuracy of some neuro-fuzzy approaches

A lot of methods have been proposed in the Literature for designing fuzzy s ystems from input-output data (the so-called neuro-fuzzy methods), but very little was done to analyze the performance of the methods from a rigorous mathematical point of view, In this paper, we establish approximation bound s for two of these methods-the table look-up scheme proposed in [15] and th e clustering method studied in [11], [13], We derive detailed formulas of t he error bounds between the nonlinear function to be approximated and the f uzzy systems designed using the methods based on input-output data, These e rror bounds show explicitly how the parameters in the two methods influence their approximation capability. We also propose modified versions for the two methods such that the designed fuzzy systems are well-defined over the whole input domain.