摘要
针对股票时间序列的非线性特点,结合混沌理论和神经网络理论,提出了基于混沌理论的股票价格神经网络预测方法。同时利用重构相空间的嵌入维数确定神经网络的结构,对实际的股票时间序列预测结果表明,该方法能有效地进行短期预测,在股票时间序列预测中有广泛的实用价值。
A method of stock price prediction based on chaos theory is presented by hypothesis of stock time series being non- linear and by taking advantages of BP neural network and chaos theory. Meanwhile, structures of neural network are determined by embedding dimension of phase space reconstruct. Predicting results for practical stock time series show that the method is able to do short- term prediction effectively, thus it can be widely used in stock time series prediction.
出处
《航空计算技术》
2009年第1期18-21,共4页
Aeronautical Computing Technique
基金
湖南省高等学校科学研究项目(08C249)
