A Neural Fuzzy System for Soft Computing

Abstract

An innovative neural fuzzy system is considered for soft computing in design. A neural tree structure is considered with nodes of neuronal type, where Gaussian function plays the role of membership function. The total tree structure effectively works as a fuzzy logic system having system inputs and outputs. In the system, as result of special provisions, the locations of the Gaussian membership functions of non-terminal nodes happen to be unity, so that the system has several desirable features; it represents a fuzzy model maintaining the transparency and effectiveness while dealing with complexity. The research is described in detail and its outstanding merits are pointed out in a framework having transparent fuzzy modeling properties and addressing complexity issues at the same time. A demonstrative application of the model is presented from a demonstrative simple architectural design exercise and the favorable performance for similar applications is highlighted.