ABSTRACT
Various sorts of data analysis are performed in a great number of areas, such as biology, medicine, economics, finance, and quality control. However, statistical expertise is a scarce and costly resource; many of these analyses are therefore performed by users with limited statistical training. In the process of knowledge acquisition and validation of the expert, a constant side-benefit of expert system development has been that new insights into the domain of expertise have arisen. Among the most intriguing characteristics of the domain of expertise in CEST are the various sorts of knowledge that are involved in the analysis of a time series. The CEST system can be thought of as a collection of small expert systems. The CEST system is to be used as a consultant; in this respect, its surface architecture is expected to resemble a DSS more than a classic automatic classification expert system.
