摘要
对价格、成交量特征样本序列分别进行小波包分解,将其小波包系数单支重构能量值共同作为样本波动的本质表征,结合遗传神经网络进行股票价格波动预测。对沪市股票上海汽车(600104)、宝钢股份(600019)进行实证研究,结果表明,该模型具有收敛速度快和预测精度高的特点。
By using wavelet pocket to analyze characteristic time series of stock price and volume, the stock price forecasting model is presented which is based on the feature vectors that reflect the energy change of wavelet packet reconstruction in several decomposition scale ranges and the genetic algorithm integrated with the back propagation algorithm. The result of two case studies of 600104 and 600019 in Shanghai Stock Exchange shows that the method converges fast and is effective.
出处
《天津大学学报（社会科学版）》
2004年第4期307-310,共4页
Journal of Tianjin University：Social Sciences
关键词
小波包分析
遗传算法
误差反向传播算法
神经网络
analysis of wavelet packet
genetic algorithm
error back propagation
artificial neural network
