Factor selection of a fuzzy system

In daily life, we normally describe our concepts and problems in linguistic terms. Due to the vagueness of our natural languages, a classical approach is unable to fully capture the properties (factors) of such concepts and problems, and hence can not provide decision makers complete information for making an appropriate decision. Therefore, in this paper, we first classify the general fuzzy data of a problem which are presented by human linguistic terms into different categories and based on their properties, each of them is described by an appropriate fuzzy set. Then, by investigating the properties of a problem as factors of a system, a fuzzy multi-objective linear programming (FMOLP) model is proposed from the viewpoint of evidence theory and information theory to measure the uncertainty of a fuzzy problem. A learning procedure is also designed to inquire the complete information according to the required level of sufficiency, /spl alpha/. Finally, an example of a mobile phone service (MPS) is presented to show the proposed model.