Dynamic information extraction for the big data
Xuebo Jin, Chao Dou · 2016
For the practical big time series data, it is the important step to eliminate the noise and get the dynamic information of the time series data. For the series data, the extraction model and the transform model are given in this paper, and the sampling interval of the models is discussed to guarantee the estimated system convergence. The dynamic information is analyzed according to the estimated dynamic characteristic. The experiment results show that for the time series data, the dynamic information extracted by Kalman filter is feasible, and can be used to analyze the trend characteristic effectively.