Abstract
The paper is concerned with an experience-consistent development of fuzzy rule-based systems. This design of such fuzzy models involves some locally available data and then reconciles the constructed model with some previously acquired domain knowledge. This type of domain knowledge is captured in the format of several rule-based models constructed on a basis of some auxiliary data sets. To emphasize the nature of modeling being guided by this reconciliation mechanism, we refer to the resulting fuzzy model as experience -consistent identification. By forming a certain extended form of the optimized performance index, it is shown that the domain knowledge captured by the individual rule-based models play a similar role as a regularization component typically encountered in identification problems. We will show that a level of achieved experience-driven consistency can be quantified through fuzzy sets (fuzzy numbers) of the parameters of the local models standing in the conclusion parts of the rules. Experimental results involve both synthetic low-dimensional data and selected data coming from data available on the Web.