DocumentCode
1748001
Title
Low complexity (turbo) classifiers in high dimensional feature spaces
Author
Tapia, Elizabeth ; González, José C.
Author_Institution
Dept. of Telematics Eng., Tech. Univ. of Madrid, Spain
fYear
2001
fDate
2001
Firstpage
41
Abstract
The hypothesis boosting concept can be understood as a kind of divide and conquer strategy for the design of low complexity classifiers. The aim of this paper is to show the feasibility of boosting algorithms in high dimension feature spaces (HDFS). A recursive learning model inspired in the design of recursive error correcting codes is proposed, with the main focus on the binary classification problem
Keywords
computational complexity; divide and conquer methods; error correction codes; information theory; learning (artificial intelligence); pattern classification; turbo codes; binary classification problem; boosting algorithms; divide and conquer strategy; high dimensional feature spaces; hypothesis boosting concept; low complexity classifiers; recursive error correcting codes; recursive learning mode; turbo classifiers; Boosting; Decoding; Error correction codes; Filters; Intelligent systems; Joining processes; Parity check codes; Strontium; Systems engineering and theory; Telematics;
fLanguage
English
Publisher
ieee
Conference_Titel
Information Theory, 2001. Proceedings. 2001 IEEE International Symposium on
Conference_Location
Washington, DC
Print_ISBN
0-7803-7123-2
Type
conf
DOI
10.1109/ISIT.2001.935904
Filename
935904
Link To Document