DocumentCode :
1118799
Title :
A Nonparametric Two-Dimensional Display for Classification
Author :
Fukunaga, Keinosuke ; Mantock, James M.
Author_Institution :
Department of Electrical Engineering, Purdue University, West Lafayette, IN 47907.
Issue :
4
fYear :
1982
fDate :
7/1/1982 12:00:00 AM
Firstpage :
427
Lastpage :
436
Abstract :
A two-dimensional display whose coordinates are related to the distance to the kth-nearest neighbor of each class is presented. Applications of the display to minimum error, minimum cost, minimax, and Neyman-Pearson type classifier designs are given. The display is shown to present risk information in a manner that easily allows the specification of reject regions. Two methods of error estimation using the display, an error counting technique and a risk averaging method, are detailed. It is shown that the classifiers that result are generalizations of the standard k-NN majority vote classifier. As a result of the properties of the display, classifiers can be readily evaluated and modified. In addition, a condensing algorithm that preserves the nearest neighbor error count of any preclassified data set is described. The display is used to graphically illustrate the distance relationships that are central to the algorithm.
Keywords :
Aerospace engineering; Costs; Error analysis; Feature extraction; Humans; Minimax techniques; Nearest neighbor searches; Pattern analysis; Two dimensional displays; Voting; Condensing algorithms; Neyman-Pearson classification; data reduction; dimensionality reduction; distance weighting; error estimation; interactive displays; minimax classification; nonparametric classifiers; reject regions;
fLanguage :
English
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on
Publisher :
ieee
ISSN :
0162-8828
Type :
jour
DOI :
10.1109/TPAMI.1982.4767276
Filename :
4767276
Link To Document :
بازگشت