Diversity and unity of uncertainty theories

摘要

Summary form only given. Serious study of the broad concept of uncertainty is an important ingredient of the emerging multidisciplinary area of intelligent systems. In this area, intelligent systems are defined as human-made systems that are capable of achieving highly complex tasks in human-like, intelligent way. The qualifier "human-like" in this definition is important since it distinguishes the area of intelligent systems from the current mainstream of artificial intelligence. In the area of intelligent systems, the human mind is viewed as a role model and the aim is to understand and emulate its various cognitive capabilities. One of them is the remarkable capability of human beings to act on perceptions, with all the imprecision and uncertainty attached to them, and to approximate perceptions in natural language. In order to emulate this capability, we need to develop much deeper understanding of the many facets of uncertainty than we possess at this time.