Title :
Tree classifier in spectral space
Author :
He, Ping ; Xu, Xiao-huax ; Chen, Ling
Author_Institution :
Dept. of Comput. Sci. & Eng., Nanjing Univ. of Aeronaut. & Astronaut., Nanjing, China
Abstract :
This paper proposes a novel nonlinear decision tree algorithm SSDT, spectral space decision tree. SSDT adopts spectral space transformation to extract the cluster information of data, employs decision tree to discover the decision boundary, and classifies test data with consistent mapping principle. Experimental results show that SSDT can produce higher classification accuracy and better generalization ability than the traditional decision tree algorithms.
Keywords :
decision trees; information retrieval; pattern classification; cluster information extraction; consistent mapping principle; data classification; nonlinear decision tree algorithm; spectral space decision tree; spectral space transformation; tree classifier; Classification tree analysis; Clustering algorithms; Computer science; Cybernetics; Data mining; Decision trees; Machine learning; Partitioning algorithms; Testing; Training data; Nonlinear tree classifier; Spectral space;
Conference_Titel :
Machine Learning and Cybernetics, 2009 International Conference on
Conference_Location :
Baoding
Print_ISBN :
978-1-4244-3702-3
Electronic_ISBN :
978-1-4244-3703-0
DOI :
10.1109/ICMLC.2009.5212576
