【作者】 梁循； 陈华； 杨健； 曾月卿；

【Author】 LIANG Xun ,CHEN Hua ,YANG Jian ,ZENG Yue-qing ( Institute of Computer Science and Technology,Peking University,Beijing 100871 .China; Department of Management Science,Stanford University,Palo Alto.CA 95035,USA)

【机构】 北京大学计算机所；

【摘要】 影响股市价格波动的因素有很多,本文从互联网信息量角度进行讨论。在一般情况下,当有较少的股市信息时,股市相对平静,股价变动也常常较小;当有较多的股市信息时,股市相对波动,股价变动常常也较大。互联网股市信息量的较大变化常常是该公司有特殊事件发生的反映,而股价波动必然是一种连带反应。本文首先从互联网获取金融信息,然后对互联网信息量进行了预处理,接着借助神经网络的学习功能,完成了对股市信息量和股市价格波动的关联学习,最后将神经网络预测的结果以图形的方式显示给投资者,帮助投资者决策。更多还原

【Abstract】 There are many factors influencing the market. This paper addresses this issue from the angle of the a-mount of the Internet information. In general,if there is less stock information,the stock price movement is often placid,and if there are lots of stock information,the stock price movement is often oscillatory. The significant increases of the amount of stock information often reflect the special events occurred in the corresponding firm, and the significant changes of stock prices simply respond to the special events. In this paper, the financial information is firstly harvested from the Internet. Then,the pre -processing is conducted. Third,based on neural networks the association is learned between the amount of stock information and stock prices. Finally,the mining result hinted by the trained neural network is demonstrated through a graphic window, helping the decisions of investors.更多还原