基于遗传神经网络算法的股票预测研究

发布时间：2018-04-30 17:03

本文选题：股市预测 + 神经网络　；参考：《兰州大学》2013年硕士论文

【摘要】：股票交易市场的波动与投资者息息相关。股市的预测研究具有很强的理论和实际意义。传统的预测方法一般是对股市进行定性和长时间范围内的预测,存在较大局限性。现在,以神经网络为代表的智能方法,由于良好的学习能力、容错性等特点,成为股市预测中较为成熟和使用较广的一种方法。本文即在此背景下,对神经网络的方法进行了介绍。基于神经网络存在的一些缺点,研究了利用遗传算法对神经网络的权值和阈值进行优化,以提高预测的速度和精度。采用上证50指数进行了实证分析。把指数价格前一天的收盘价和当天的开盘价作为输入样本,预测当天的收盘价。结果表明,神经网络经过遗传算法优化后,预测结果比原先单纯使用神经网络方法有所提高,结果令人满意。但是,输入量的选择是否合理、神经网络和遗传算法中参数确定并未有明确理论指导等问题依然有待解决,这些也都是在运用智能方法进行股市预测中值得进一步探讨的问题。
[Abstract]:The volatility of the stock market is closely related to investors. The research of stock market prediction has strong theoretical and practical significance. The traditional forecasting method is to predict the stock market qualitatively and within a long period of time, which has some limitations. Now the intelligent method represented by neural network has become a mature and widely used method in stock market forecasting because of its good learning ability and fault tolerance. In this context, the method of neural network is introduced in this paper. Based on the shortcomings of neural networks, the genetic algorithm is used to optimize the weights and thresholds of neural networks in order to improve the speed and accuracy of prediction. Using the Shanghai Stock Exchange 50 index for empirical analysis. The closing price of the previous day and the opening price of the day were used as input samples to forecast the closing price of the day. The results show that the prediction results of the neural network are better than those of the original neural network method after genetic algorithm optimization, and the results are satisfactory. However, whether the selection of input is reasonable or not, and whether the parameters in neural network and genetic algorithm are not clear theoretical guidance are still to be solved, which are also worthy of further discussion in the use of intelligent methods for stock market prediction.
【学位授予单位】：兰州大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP18;F830.91

【参考文献】