Application of Wavelets and Hilbert- Huang Transform On Predicting Stock Forecasting

• Main Authors: Ru-ling Tsai, 蔡茹鈴
• Other Authors: Kuei-Fang Chang
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/12788362178257669543

碩士 === 逢甲大學 === 應用數學所 === 96 === In this paper, we choose the increase stock prices of MediaTek and decrease stock prices of Everlight Chemical Industrial Corporation as experimental samples. After applying Discrete Wavelet Transform and HHT into processing, we use first-order auto-regression model to predict next ten indexes step by step, comparing two methods by evaluating their MSE; in addition, we use envelope of EMD to produce upper and lower envelope, between which is the region called a feasible band. It is important that feasible band is contained in confidence band. Finally, we use Kolmogorov-Smirnov Test to check the normal distribution and see whether the mean of the residue is zero or not; then, in terms of Analysis of Variance, we point out that the variances between real stock prices and observed prices are close.