Application of gradient descent and genetic algorithms to the trapezoidal fuzzy set based approximate SAM

Abstract

Parameter learning and Optimal learning, as well to all other knowledge presentative models, have an essential sense to the quality and effectivity in approximation of SAM. This paper presents not only an applied research on gradient descent algorithm and genetic algorithm to the parameter learning process and the optimal learning process of trapezoidal fuzzy set based SAM system, but also, an application of the research to the time-series prediction.