An interpolative fuzzy inference using least square principle by means of /spl beta/-function and high order polynomials

Many researchers have been interested in approximation properties of fuzzy logic systems (FLS), which like neural networks, can be seen as approximation schemes. Almost all of them tackled Mamdani fuzzy model, which was shown to have many interesting features. This paper aims to present an alternative for traditional inference mechanisms and CRI method. The most attractive advantage of this new method is its higher robustness with respect to changes in rule base and ability to operate when latter is sparse. In this paper interpolation with high order polynomials and /spl beta/-function is reported.