Title :
A weighted minimum distance classifier for pattern recognition
Author :
Lin, H. ; Venetsanopoulos, A.N.
Author_Institution :
Dept. of Electr. Eng., Toronto Univ., Ont., Canada
Abstract :
This paper presents a new weighted minimum distance classifier which uses the discriminately power and variance of features. The weights increase the interclass separability while they decrease the intraclass dissimilarity. Two examples are given to show the effectiveness of the method
Keywords :
matrix algebra; pattern recognition; interclass separability; intraclass dissimilarity; pattern recognition; weighted minimum distance classifier; Cities and towns; Density functional theory; Density measurement; Equations; Euclidean distance; Multi-layer neural network; Neural networks; Pattern recognition; Power engineering and energy;
Conference_Titel :
Electrical and Computer Engineering, 1993. Canadian Conference on
Conference_Location :
Vancouver, BC
Print_ISBN :
0-7803-2416-1
DOI :
10.1109/CCECE.1993.332440
