The inverse Castelzho mapping in neurofuzzy models
Abstract
Neurofuzzy model construction algorithm for nonlinear dynamic system identification is considered using Bernstein polynomials as fuzzy membership functions. A new analytical method for the inverse Castelzho mapping is proposed to determine barycentric coordinates in bivariate function approximation by the Bezier surface. Examples of using new approach for the real problems of identification some equations in the simple macroeconometric model of Ukraine’s economy are shown.Downloads
Issue
Section
New methods in system analysis, computer science and theory of decision making