基于FAM-ELM股票价格预测研究

发布时间：2018-01-04 20:05

本文关键词：基于FAM-ELM股票价格预测研究　出处：《计算机仿真》2014年08期 　论文类型：期刊论文

【摘要】：研究股票价格准确预测问题。股票价格预测是股票交易者最关心的问题,直接影响着股票交易者的收益。由于股票受经济发展的影响,价格波动较大,在股票价格预测中采用传统神经网络方法存在训练速度慢,易陷入局部极小值,隐含层节点数人为指定等问题,导致泛化能力受到影响,预测不准。为了提高股票价格预测的精度,提出基于因子分析法的极限学习机股票价格预测模型。首先使用因子分析法综合股票价格影响指标;接着使用隐含层神经元数量寻优算法搜索最优隐含层神经元数量值;然后使用极限学习机对综合后的股票价格影响指标进行学习,建立股票价格预测模型;最后通过实验对模型性能进行测试。试验结果证明,基于因子分析法的极限学习机提高了股票价格的预测精度和运行效率。
[Abstract]:Research on accurate prediction of stock price. Stock price prediction is the most concerned issue of stock traders, which directly affects the income of stock traders. Because of the impact of economic development, stock prices fluctuate greatly. In the stock price prediction, the traditional neural network method has some problems, such as slow training speed, easy to fall into local minimum value, and the number of hidden layer nodes is designated, which results in the influence of generalization ability. In order to improve the accuracy of stock price prediction, a model of extreme learning machine stock price prediction based on factor analysis is proposed. Firstly, factor analysis method is used to synthesize stock price influence index. Then the optimal number of hidden layer neurons is searched by using the optimal algorithm of the number of hidden layer neurons. Then we use the extreme learning machine to study the comprehensive stock price influence index and establish the stock price prediction model. Finally, the performance of the model is tested by experiments. The experimental results show that the ultimate learning machine based on factor analysis can improve the precision and efficiency of stock price prediction.
【作者单位】：西安邮电大学管理工程学院;
【基金】：陕西省自然科学基金(2012GQ8050) 陕西省教育厅专项科研计划项目(13JK0403) 西安邮电大学中青年基金(104-0410)
【分类号】：F830.91;TP18
【正文快照】： 1引言股票不但是国民经济的晴雨表,同时也是市民热衷的投资工具,其价格波动的预测方法一直以来都是金融领域研究的焦点问题之一[1]。目前,国内外对于股票价格波动进行预测的模型很多,按照建模理论的不同,可将这些预测模型划分为两个大类:一类是以统计原理为基础的传统型波动率

【参考文献】