DocumentCode
401665
Title
Research of time series pattern finding based on artificial neural network
Author
Peng, Yu-Qing ; Zhang, Yu ; Tian, Hai-Shan
Author_Institution
Sch. of Comput. Sci. & Eng., Hebei Univ. & Technol., Tianjin, China
Volume
3
fYear
2003
fDate
2-5 Nov. 2003
Firstpage
1385
Abstract
In this paper, a new way of time-series pattern finding is presented. It fits the time-series curve with sequence of straight lines and the line´s arctg is obtain by a piecewise linear approximation. The arctg is regarded as input sample, and artificial neural network and hierarchical processing technology is applied. The experiments prove that the classification accuracy based on ANN is improved.
Keywords
data mining; hierarchical systems; neural nets; pattern classification; piecewise linear techniques; time series; artificial neural network; data mining; hierarchical processing technology; pattern classification; piecewise linear approximation; time series pattern finding; Artificial neural networks; Computer science; Curve fitting; Data mining; Delay effects; Electronic mail; Finishing; Frequency domain analysis; Pattern matching; Piecewise linear approximation;
fLanguage
English
Publisher
ieee
Conference_Titel
Machine Learning and Cybernetics, 2003 International Conference on
Print_ISBN
0-7803-8131-9
Type
conf
DOI
10.1109/ICMLC.2003.1259708
Filename
1259708
Link To Document