OWA Aggregation on an Interval Argument

Abstract

The OWA operator provides a parameterized family of averaging operators. It has as its argument a finite collection of values. This chapter extends the applicability of the OWA operator to the situation in which the argument is a continuous interval. It presents the role of the attitudinal character of the OWA weighting vector in this aggregation process considering the extension of the continuous interval argument OWA operator to the more general case in which the argument values have importance weights. In many applications involving random variables, one makes use of a unique scalar value as a replacement for the more complex probability distribution. This simplification allows performing many tasks, such as comparisons, that may be very difficult if not impossible using probability distributions. The expected value is most often used to provide the representative scalar value.