Unlike low frequency data, high frequency data has the calendareffects and long memory volatility.
与低频数据不同，高频数据通常具有“日历效应”和波动长记忆性。
2
The paper proposes application of Wavelet Neural Network in high-frequency time series calendareffects' study. At last, the paper proves that WNN is better than classical FFF regression.
Calendareffects mean that market returns associate with the specific transaction date in stock market, there re two important forms: day of the week effects and month of the year effects.