论文基于BP神经网络的上证股票指数预测.doc

基于BP神经网络的上证股票指数预测 Shanghai Stock Index Prediction with Neural Networks 内容提要 证券市场作为高风险高收益的投资领域一直倍受投资者的关注，如何能够准确分析和预测股票价格以便获取丰厚的收益一直受到人们的关注。于是各种股票价格分析预测和方法应运而生，如何建立一个成功率比较高的预测理论和模型是多年来许多学者一直研究的内容。 随着证券市场混沌和分形理论的逐步确立，人们开始利用神经网络对证券市场的变动加以预测。神经网络是一种重要的人工智能技术，它的研究开始于20世纪40年代，近年来，神经网络任意连续映射的逼近能力学习理论以及动态网络的稳定性分析都已取得了丰硕的成果；在应用上也迅速扩展到许多重要的领域涉及模式识别与图象处理、控制与优化、ATM网络中呼叫接纳的识别与控制、导航多媒体处理系统等等。神经网络的自适应学习非线性映射强的特点非常适合应用于经济领域的信息处理以及分析时间序列。 BP（ Back Propagation） 网络是一种被广泛运用的神经网络。它的核心是BP算法，一种对于多基本子系统构成的大系统进行微商计算的严格而有效的方法，结构简单，算法成熟。与传统统计回归方法相比，神经网络不仅能够学习训练集的例子，且能从训练集中提炼出某种一般性原理、规律，具有很强 的非线性函数拟合特性, 这对于预测短周期内股指波动有较强的适用性。 本文尝试利用基于BP算法的三层向前神经网络对上海交易所上证指数进行了预测，并针对BP网络原形的一些缺点和不足，对原有的预测方法作出了一些改进。在实际预测中取得了良好的效果。 关键词：指数预测；神经网络； BP算法 Abstract To be an investment region of high risk and high profit, Stock market attracts many investors’ attentions all along. How to obtain profit through analyzing and forecasting the share price accurately is attracting the people’s attentions. With the developing of Fractal Market Hypothesis (FMH) theory, people have tried to forecast the change of the Stock market. The neural network is an important technology in the filed of AI, which was developed in 1940’s. In recent years, the theory about the approximation of random continuous maps by neural network and the analyzing of the stability of a dynamic network have been used in many fields and gained great achievement. BP (Back Propagation) is a neural network which is adopted widely. The core is the BP arithmetic, a strict and effective method to derivative problem for system based on multi-subsystem, which has simple configuration and mature arithmetic. To compare with the traditional statistical regress method, BP network can not only study the example of training set, but also abstract some general theory and rule. It has strong characteristic of approximation of non-linear functions, which is much fit for stock index analyzed and predicted in a short-term. This article tries to use a neural network on