Article ID Journal Published Year Pages File Type
403425 Knowledge-Based Systems 2016 8 Pages PDF
Abstract

We introduce the OWA operator and note that it provides a parameterized class of aggregation operators. Here the parameterization is accomplished by the choice of the characterizing OWA weights, different characterizing weights results in different aggregation imperatives. We discuss various ways of providing these characterizing OWA weights. Most notable among these are the use of a vector containing the prescribed weights and the use of a function called the weight generating function from which the characterizing can be extracted. In many applications we are faced with situations in which the arguments being aggregated have different importances. This raises the issue of appropriately combining the individual argument weights with the characterizing weights of the operator to obtain operational weights to be used in the actual aggregation. Our goal here is looking at this issue under different methods of specification of the characterizing weights.

Related Topics
Physical Sciences and Engineering Computer Science Artificial Intelligence
Authors
, ,