DocumentCode :
2208611
Title :
Adaptive Distances on Sets of Vectors
Author :
Woznica, Adam ; Kalousis, Alexandros
Author_Institution :
Comput. Sci. Dept., Univ. of Geneva, Carouge, Switzerland
fYear :
2010
fDate :
13-17 Dec. 2010
Firstpage :
579
Lastpage :
588
Abstract :
Recently, there has been a growing interest in learning distances directly from training data. While the previous works focused mainly on adapting distance measures over vectorial data, it is a well-known fact that many real-world data could not be easily represented as fixed length tuples of constants. In this paper we address this limitation and propose a novel class of distance learning techniques for learning problems in which instances are set of vectors, examples of such problems include, among others, automatic image annotation and graph classification. We investigate the behavior of the adaptive set distances on a number of artificial and real-world problems and demonstrate that they improve over the standard set distances.
Keywords :
learning (artificial intelligence); set theory; vectors; k-nearest neighbor algorithm; learning distances; training data; tuples; vector; vectorial data; adaptive distances; complex objects; distance learning; graphs; sets;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Data Mining (ICDM), 2010 IEEE 10th International Conference on
Conference_Location :
Sydney, NSW
ISSN :
1550-4786
Print_ISBN :
978-1-4244-9131-5
Electronic_ISBN :
1550-4786
Type :
conf
DOI :
10.1109/ICDM.2010.45
Filename :
5694012
Link To Document :
بازگشت