摘要股票作为市场经济的重要组成部分,是金融市场的报警器。股票是从作为市场经济相当重要的部分之一,在金融领域有着不可替代的地位,也将影响金融市场健康发展和稳定。因此对股票市场的关注程度一直火热,而股票股价的预测也成为一个热门的课题。本文从股票市场的重要性开始叙述,分析几种基本的预测方法,像技术分析法,时间序列法。通过资料收集发现神经网络法对于股票的预测有一定的实用性。第二章介绍了人工神经网络的起源,还有人工神经网络模型的计算思路。第三章开始着重对神经网络中的BP神经网络进行介绍,研究了学习的算法,发现BP神经网络的局限之处,阐述了优化神经网络几个思路,并对改进后的算法进行可行性分析,总结出了具体的分析步骤。第四章着重就改进后的BP神经网络进行实例分析。从分析的步骤,MATLAB实现等方面入手,进行仿真训练,将得到的结果和标准的神经网络的算法进行对比,得出结论。第五章就实验的不足之处提出了学习的地方,还有对BP神经网络的展望。48453
Abstract As an important part of market economy, stock is alarm in financial markets. Stock from a quite important part of market economy, has an irreplaceable position in the financial sector, will also affect the healthy development of financial market and stable. So the stock market has been on hot, and the stock price forecasting has become a hot topic. This paper, starting from the importance of the stock market prediction approach for the analysis of several basic, like technical analysis, time series method. Through data collection found that neural network for the prediction of the stock has a certain practicality. The second chapter introduces the origin of the artificial neural network and artificial neural network model of computation. Began to focus on the third chapter to introduce in neural network, BP neural network, studied the learning algorithm of discovering the limits of the BP neural network, this paper expounds the optimization of neural network to a few ideas, and carries on the feasibility analysis of the improved algorithm, the concrete steps of analysis were summarized. The fourth chapter focuses on the improved BP neural network is analyzed. From the analysis steps, MATLAB, via the aspects of the simulation training, will get the results were compared with the standard of the neural network algorithm, the conclusion. The deficiencies in the fifth chapter is the experiment put forward learning place, along with the outlook of the BP neural network.
毕业论文关键词:神经网络; 预测; 股价; 仿真
Keyword: neural network; forecasting; stock price; simulation
目 录
一、引言 1
1.1课题的研究背景 1
1.2传统的股票预测方法 1
1.3研究目的及意义 2
二、神经网络概述 2
2.1神经网络来源 2
2.2人工神经网络 2
2.3神经网络拓扑结构 4
三.BP(Back-Propagation)神经网络 4
3.1 BP神经元模型 4
3.2 BP神经网络的应用 5
3.3 BP神经网络的学习算法 5
3.4 BP神经网络的不足和局限性 8
3.5 BP算法的优化措施 9
3.6 BP神经网络应用于股价预测的可行性分析