| Wavelet analysis (WA) is a branch of applied mathematics developed by Y. Meyer, S. Mallat and I. Daubechies since 1986. It now is extensively concerned and has widely applications. It is a milestone during the development of Fourier, but Fourier cannot analyze partial signal in time field. Because of "self-adaptable" and "focus-adjustable", WA plays an important role in processing the non-stationary signal. Multiresolution analysis expresses the reduced space based on these characteristics. This paper is mainly focused on the application of WA in the data processing in financial market and the analysis of economic time series.First of all, the period of future market is studied by wavelet decomposition and multiresolution analysis, which map different price signals to specific frequencies. Secondly, the soybean indices from Dalian Commodity Exchange and CBOT market are concisely compared, after using filter to denoise the financial time series, which can sufficiently preserve the characteristics of original signals. Time series decomposition and reconstruction model is established by analyzing the different decomposition parts of the time series, which is based on wavelet analysis and ARMA model. The results of forecasting prices in agriculture future market by this model are more accurate than that by traditional ARMA model. At last, the validity of Chinese future market is discussed based on these results. |
PDF Full Text Request