基于模糊K线序列比对的股市技术分析模型  

TECHNICAL ANALYSIS MODEL FOR STOCK MARKET BASED ON FUZZY CANDLESTICKS SEQUENCE ALIGNMENT

作　　者：徐信喆[1] 

机构地区：[1]上海交通大学计算机科学与工程系生物计算与生物信息实验室,上海210030

出　　处：《计算机应用与软件》2010年第9期28-32,48,共6页Computer Applications and Software

基　　金："985工程"二期建设项目(7222103003)

摘　　要：提出一种新的股市技术分析模型,该模型利用模糊逻辑理论及生物序列比对方法的思想,加以改进,应用于传统的K线图理论,将单位时间的开盘、收盘、最高、最低价格编码成模糊K线图,通过模糊序列比对,来对K线图进行模式识别。以上海和深圳自1990年开市以来到2006年的所有数据作为比对数据库,以2007年、2008年的数据作为测试对象,对上海和深圳股票市场中的部分股票中的K线图模式及未来趋势作了分析和预测,取得了可观的结果。统计结果表明,一些K线模式序列的未来趋势分布具有涨跌信号预警的功能;而另一些模式则为平凡的序列匹配,它们未来趋势的分布不具有预警的功能。In this paper a new technical analysis model for stock market is proposed, which integrates fuzzy logical theory and biological se- quence alignment together and applies the improved approach in traditional Japanese candlestick chart theory. Based on fuzzy sequence align- ment of fuzzy candlestick chart encoded with the open, close, high and low prices over time, the pattern recognition of candlestick chart a- chieves. The data from both Shanghai Stock Exchange and Shenzhen Stock Exchange dating from 1990-2006 are used as comparing databases and the data in 2007 and 2008 are used as testing objects, the patterns of candlestick charts and their future development trends of part stocks collected from the two institutions are analysed and predicted. The results derived are very promising. Statistical results indicate that distribu- tion of future trend of some candlestick sequence patterns indicate a very clear bullish or bearish signal whereas some other patterns are mere result of trivial sequence matches and their distribution of future trend can not function as a predictive signal.

关 键 词：K线图 模糊K线序列 模糊逻辑 序列比对 模式识别 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]