小波网络在深圳股市应用的研究  

Study on Applying of Wavelet Transform and Neural Networks in Shenzhen Stock Market

作　　者：金玲玲[1] 汪刘一[2] 

机构地区：[1]华南农业大学理学院,广东广州510642 [2]华南农业大学工程学院,广东广州510642

出　　处：《华南农业大学学报》2003年第3期82-84,共3页Journal of South China Agricultural University

基　　金：华南农业大学校长基金资助项目(4900-k02021)

摘　　要：采用传统的人工神经网络模型对深圳证券成份指数进行模拟预测,在此基础上,进一步采用小波函数结合神经网络形成的小波网络对其进行拟合和预测,并对两种预测方法得到的结论误差进行分析、比较.结果显示,小波网络比单纯的神经网络模型预测精度高11 1595.In the first approach, the conventional artificial neural networks(fuzzy perce ption) were used to approximate and predict the Shenzhen composite index. Based on this result, wavelet network were utilized, which is a hybrid of wavelet transform and neural network , to carry the same task. Comparing the prediction errors from both approaches clearly show that wavelet network offers very satisfied results.

关 键 词：小波网络 深圳股市 应用 小波分析 神经网络 股价 股票价格指数 

分 类 号：F832.5[经济管理—金融学]