Understanding the level of uncertainty associated with a prediction is valuable in determining its utility in decision making. One measure of information is Yager's notion of specificity. Yager views specificity as the degree to which a possibility distribution points to a single element in the universe of discourse (U). Specificity in relation to U may complicate its utility in the optimization of fuzzy models in their linguistic space. An increase in granularity is useful to measure the amount of information contained in a possibility distribution in relation to fuzzy sets as opposed to U. This abstracted view of specificity motivates the need for a more generalized version of specificity, denoted Linguistic Specificity (SpL), where alternatives are measured in relation to the linguistic terms. Such a generalization is useful in support of automating decisions in a fuzzy domain. Results of the linguistic specificity measure are illustrated using an automobile fuel consumption example.
|Original language||English (US)|
|Number of pages||5|
|Journal||Annual Conference of the North American Fuzzy Information Processing Society - NAFIPS|
|State||Published - 1999|
All Science Journal Classification (ASJC) codes
- Computer Science(all)