Linguistic Nearest Prototype (LNP)

Linguistic Nearest Prototype (LNP) is the extended version of the nearest prototype based on the extension principle and the decomposition theory. The input vectors are vectors of fuzzy subsets. The prototype vectors are vectors of fuzzy subsets as well. The LNP classifies an input vectors into the same class as the closest prototype vector. The fuzzy city block, i.e., an extended version of the city block distance is used to determine the closeness. After the fuzzy distances are computed, they are ranked by any ranking method, e.g., centroid [Yager78], indexing by mean interval [Gonzalez90], indexing by area [Choobineh93], and etc.

Example

input vectors prototype vectors

main page