基于GD-FNN的股票市场泡沫模型
发布时间：2018-12-10 07:22
【摘要】：针对股票市场内部结构复杂性和外部因素多变性,构建一种基于预测的股票市场泡沫模型.以上证指数为研究对象,在价格和成交量的基础上,将与股票市场密切相关的宏观经济指标引入泡沫模型指标体系,并对指标体系中各变量之间长期均衡关系和因果关系进行数量分析.在此指标体系下,构建向量自回归模型(VAR)模型衡量股票市场基础价值,并据此分析宏观经济指标对市场的影响;同时构建基于椭圆基函数且能够动态调整网络结构的广义动态模糊神经网络模型(GD-FNN)对上证指数进行拟合预测作为股票市场的市场价值,并通过GD-FNN模型提取的模糊规则对股票非线性系统运行模式进行分析.最后,根据预测的股票市场市场价值与基础价值之间的偏差计算泡沫度,并提出相应的预警策略.
[Abstract]:Aiming at the complexity of internal structure and the variability of external factors in stock market, a foresight model of stock market bubble is constructed. Taking the Shanghai Stock Exchange Index as the research object, on the basis of the price and the trading volume, the macro-economic index closely related to the stock market is introduced into the index system of the bubble model. The long-term equilibrium and causality among the variables in the index system are analyzed quantitatively. In this index system, we construct the vector autoregressive model (VAR) to measure the fundamental value of stock market, and then analyze the influence of macroeconomic indicators on the market. At the same time, the generalized dynamic fuzzy neural network model (GD-FNN), which is based on elliptic basis function and can dynamically adjust the network structure, is constructed to predict Shanghai stock index as the market value of stock market. The operation mode of stock nonlinear system is analyzed by fuzzy rules extracted from GD-FNN model. Finally, the foaming degree is calculated according to the deviation between the predicted stock market value and the basic value, and the corresponding early warning strategy is put forward.
【作者单位】： 北京科技大学
【基金】：国家自然科学基金(70771008,70371057) 北京科技大学博士研究生科研基金项目
【分类号】：F224;F830.91
[Abstract]:Aiming at the complexity of internal structure and the variability of external factors in stock market, a foresight model of stock market bubble is constructed. Taking the Shanghai Stock Exchange Index as the research object, on the basis of the price and the trading volume, the macro-economic index closely related to the stock market is introduced into the index system of the bubble model. The long-term equilibrium and causality among the variables in the index system are analyzed quantitatively. In this index system, we construct the vector autoregressive model (VAR) to measure the fundamental value of stock market, and then analyze the influence of macroeconomic indicators on the market. At the same time, the generalized dynamic fuzzy neural network model (GD-FNN), which is based on elliptic basis function and can dynamically adjust the network structure, is constructed to predict Shanghai stock index as the market value of stock market. The operation mode of stock nonlinear system is analyzed by fuzzy rules extracted from GD-FNN model. Finally, the foaming degree is calculated according to the deviation between the predicted stock market value and the basic value, and the corresponding early warning strategy is put forward.
【作者单位】： 北京科技大学
【基金】：国家自然科学基金(70771008,70371057) 北京科技大学博士研究生科研基金项目
【分类号】：F224;F830.91
【参考文献】
相关期刊论文 前10条
1 陈莹;赵成国;李心丹;;中国证券市场泡沫测度及形成机理研究[J];复旦学报(社会科学版);2010年02期
2 汤凌冰,廖福元,罗键;模糊神经网络在股价预测中的应用[J];系统工程;2004年02期
3 周松柏;;中国证券市场泡沫问题实证研究[J];系统工程;2009年09期
4 攀登;施东晖;宋铮;;证券市场泡沫的生成机理分析——基于宝钢权证自然实验的实证研究[J];管理世界;2008年04期
5 孙彬;李铁克;张文学;;基于DFNN的金融股指预测及金融非线性系统辨识研究[J];中国管理信息化;2009年21期
6 陈永清,韩德宗;金融资产泡沫实证研究[J];经济理论与经济管理;2002年10期
7 刘澄;张均东;孙彬;;略论ANFIS在金融股指预测中的应用[J];经济问题;2009年11期
8 黄秀海;;一种新的股市泡沫计量方法[J];经济学家;2008年01期
9 刘q，
本文编号：2370190
本文链接：https://www.wllwen.com/jingjilunwen/guojijinrong/2370190.html
