基于情绪指数和神经网络的上证指数预测研究
发布时间：2018-07-28 09:28
【摘要】：与现代经济不断的发展相适应，大众的投资意识不断地发生变化，股票投资已经成为人们投资理财的重要手段，股票投资日益变成社会大众关心的热门话题，因此股票市场的质量以及繁荣程度成为了管理人员和投资者关注和研究的热点。股票投资有收益-风险成正比的特点，即预期收益越高，伴随的可能的风险也越大。因而，对股票市场预测方法的探索具有非常高的经济价值和理论意义。由于股票的价格形成机制十分复杂，有很多的外部因素都会对其产生影响，因此要准确的预测股票市场不是一项容易的任务，以往使用过的预测方法在股市预测的应用中都不能取得非常满意的效果。 本文在深入分析了行为金融学理论以及神经网络建模理论的基础上，选取交易价格和交易者情绪这两个能够反映市场信息和价格形成过程的维度对上证指数建立一个BP神经网络模型进行短期的预测，用以检验BP神经网络在股价预测方面的应用效果。在以往关于投资者情绪的研究中通常的作法是选取一个技术指标作为交易者情绪的代理变量，，本文在前人研究的基础上，应用主成分分析法构造了一个交易者情绪指数用来代表市场上交易者的情绪，进而作为预测模型的一个输入变量。同时，对上证指数进行线性模型的建模，作为对比模型来检测神经网络的预测效果。 理论分析和实验结果说明，应用神经网络对上证指数进行预测具有一定的有效性，加入构造的情绪指数能够显著的提高模型的预测精度，在股指预测领域有一定的应用价值。
[Abstract]:In keeping with the constant development of modern economy, the public's investment consciousness is constantly changing, and stock investment has become an important means for people to invest and manage money. Stock investment has increasingly become a hot topic of public concern. Therefore, the quality and prosperity of stock market have become the focus of attention and research by managers and investors. The higher the expected return, the greater the possible risk associated with stock investment. Therefore, the exploration of stock market forecasting method has high economic value and theoretical significance. Because the stock price formation mechanism is very complex, there are many external factors will have an impact on it, so it is not an easy task to accurately predict the stock market. The prediction methods used in the past can not achieve very satisfactory results in the application of stock market forecasting. Based on the in-depth analysis of behavioral finance theory and neural network modeling theory, Selecting the two dimensions which can reflect the process of market information and price formation, the transaction price and the traders' sentiment are selected to establish a BP neural network model for short-term prediction of Shanghai Stock Exchange Index. It is used to test the application effect of BP neural network in stock price prediction. In the past studies on investor sentiment, the usual method is to select a technical index as the proxy variable of traders' sentiment. The principal component analysis (PCA) is used to construct a trader's emotion index to represent the traders' emotion in the market, and then it is used as an input variable of the prediction model. At the same time, the linear model of Shanghai Stock Exchange Index is built to detect the prediction effect of neural network as a contrast model. The theoretical analysis and experimental results show that the application of neural network to the prediction of Shanghai stock index has a certain effectiveness, adding the constructed emotional index can significantly improve the prediction accuracy of the model, and has a certain application value in the field of stock index prediction.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.51;F224
本文编号：2149656
[Abstract]:In keeping with the constant development of modern economy, the public's investment consciousness is constantly changing, and stock investment has become an important means for people to invest and manage money. Stock investment has increasingly become a hot topic of public concern. Therefore, the quality and prosperity of stock market have become the focus of attention and research by managers and investors. The higher the expected return, the greater the possible risk associated with stock investment. Therefore, the exploration of stock market forecasting method has high economic value and theoretical significance. Because the stock price formation mechanism is very complex, there are many external factors will have an impact on it, so it is not an easy task to accurately predict the stock market. The prediction methods used in the past can not achieve very satisfactory results in the application of stock market forecasting. Based on the in-depth analysis of behavioral finance theory and neural network modeling theory, Selecting the two dimensions which can reflect the process of market information and price formation, the transaction price and the traders' sentiment are selected to establish a BP neural network model for short-term prediction of Shanghai Stock Exchange Index. It is used to test the application effect of BP neural network in stock price prediction. In the past studies on investor sentiment, the usual method is to select a technical index as the proxy variable of traders' sentiment. The principal component analysis (PCA) is used to construct a trader's emotion index to represent the traders' emotion in the market, and then it is used as an input variable of the prediction model. At the same time, the linear model of Shanghai Stock Exchange Index is built to detect the prediction effect of neural network as a contrast model. The theoretical analysis and experimental results show that the application of neural network to the prediction of Shanghai stock index has a certain effectiveness, adding the constructed emotional index can significantly improve the prediction accuracy of the model, and has a certain application value in the field of stock index prediction.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.51;F224
【参考文献】
相关期刊论文 前10条
1 孙延风,梁艳春,孟庆福;改进的神经网络最近邻聚类学习算法及其应用[J];吉林大学学报(信息科学版);2002年01期
2 孙丹,张秀艳;基于人工神经网络的股市预测模型[J];吉林大学学报(信息科学版);2002年04期
3 高琴;谈玲;;基于BP神经网络的股市预测模型[J];电脑知识与技术(学术交流);2007年02期
4 刘煜辉,熊鹏;资产流动性、投资者情绪与中国封闭式基金之谜[J];管理世界;2004年03期
5 李国平;;中国股票市场的可预测性研究[J];高职论丛;2006年03期
6 张立军;苑迪;;基于GA-Elman动态回归神经网络的股价预测模型研究[J];华东经济管理;2008年09期
7 黄少军;中国股票市场可预测性研究[J];华南师范大学学报(社会科学版);2004年02期
8 郝勇;;基于MATLAB神经网络工具箱的上海证券商业指数的预测分析[J];经济师;2005年12期
9 李心丹,王冀宁,傅浩;中国个体证券投资者交易行为的实证研究[J];经济研究;2002年11期
10 王美今,孙建军;中国股市收益、收益波动与投资者情绪[J];经济研究;2004年10期
本文编号：2149656
本文链接：https://www.wllwen.com/guanlilunwen/huobilw/2149656.html
