Title :
A method for data stream processing based on curve fitting
Author :
Song, Yixu ; Hu, Jing ; Yang, Xiaokui ; Fu, Jie ; Xie, Xiufen
Author_Institution :
Dept. of Comput. Sci. & Technol., Tsinghua Univ., Beijing, China
Abstract :
The sampling storage method which used in the current data stream could not respond data tendency effectively. For the problem, this paper presents a new processing method based on curve fitting. A weighted least-square principle is used to fit the cached stream data and better model description is obtained. Then the fitting results are analyzed by clustering algorithm, which serves as a classifier for polynomial fitting parameters. According to the clustering result, the appropriate window size will be given to fit the periodic stream data. Comparing the function solutions with the actual data, the different methods are adopted to store data according to the comparison result. The experimental results indicate that the proposed method has better fitting accuracy and compression ratio, could meet the requirement of data stream processing. And the data tendency could be responded effectively by the fitting results.
Keywords :
curve fitting; data compression; least squares approximations; media streaming; pattern clustering; polynomials; sampling methods; cached stream data; clustering algorithm; curve fitting; data stream processing; periodic stream data; polynomial fitting parameters; sampling storage method; weighted least square principle; Accuracy; Algorithm design and analysis; Clustering algorithms; Compression algorithms; Curve fitting; Fitting; Signal processing algorithms; clustering; curve fitting; data stream; least -square principle;
Conference_Titel :
Signal Processing Systems (ICSPS), 2010 2nd International Conference on
Conference_Location :
Dalian
Print_ISBN :
978-1-4244-6892-8
Electronic_ISBN :
978-1-4244-6893-5
DOI :
10.1109/ICSPS.2010.5555670
