基于模糊K线的金融时间序列反转模式挖掘研究

发布时间：2018-04-29 17:57

本文选题：模糊K线 + 时间序列　；参考：《湖南大学》2012年硕士论文

【摘要】：股票市场是一个受多种因素影响的庞大系统，具有非常复杂的运动规律，市场行情更是瞬息万变。股票市场的金融时间序列数据作为其综合的外在表现形式，蕴含了许多客观规律信息，如何从中挖掘出各种信息，更好地认识、掌握、并利用其规律无疑对股票投资预测、决策与风险管理活动具有重要意义。传统技术分析主要是基于经验对市场变化的趋势进行预测，以图表为手段对市场行为进行研究。由于不同的人对图表的识别差异较大，难以统计验证到底谁的分析结果最可靠，并且缺乏足够的理论支持。现代技术分析则主要是运用数学模型和数理统计的方法，试图从大量的历史数据中挖掘出市场在时间序列层面上可能隐藏的规律。但现有的研究中，，多数预测模式参数繁多且复杂，难以被一般投资者所理解，同时现有的研究主要着眼于对未来价格或趋势的预测，事实上，预测未来股票的具体价格是一个非常困难的工作，毕竟股票市场存在太多的不确定因素。市场中的投资者更关心、而且也较为可行的策略是识别市场的反转点。本文研究了一个能主动预警股价反转点的新技术分析模型，该模型将模糊逻辑理论应用于传统的K线图理论，利用K线图的开盘价、收盘价、最高价和最低价对K线形态特征进行模糊化处理，再从模糊K线图中提取出征兆序列，然后运用数据挖掘技术对征兆序列进行分类，得出模糊K线反转模式的统计特征，最后通过模糊K线反转模式实现对股价反转点的预测。本文选择在上证A股和深证A股这两个证券市场分别进行实证研究，结果表明：本文提出的模糊K线反转模式真实存在并且能够被较好的识别；根据本文提出的模糊K线反转模式得到的股价预测反转点进行模拟股票交易所获投资回报率要高于随机交易所得投资回报率。
[Abstract]:The stock market is a huge system influenced by many factors. The financial time series data of the stock market, as its comprehensive external form, contain a lot of objective law information, how to dig out all kinds of information from the stock market, better understand, master, and make use of its law to predict the stock investment without doubt. Decision-making and risk management activities are of great significance. The traditional technical analysis is mainly based on experience to predict the trend of market change, using charts as a means to study market behavior. It is difficult to verify whose analysis results are the most reliable and lack of sufficient theoretical support due to the large differences in the recognition of charts by different people. The modern technology analysis mainly uses the mathematical model and the mathematical statistics method, tries to excavate the law which the market may hide in the time series level from the massive historical data. But in the existing research, most of the prediction model parameters are various and complex, which is difficult to be understood by the general investors. At the same time, the existing research mainly focuses on the prediction of future prices or trends, in fact, Predicting the exact price of stocks in the future is a very difficult task. After all, there are too many uncertainties in the stock market. Investors in the market are more concerned, and more feasible, strategy is to identify the market reversal point. In this paper, we study a new technical analysis model that can proactively warn the reverse point of stock price. In this model, fuzzy logic theory is applied to the traditional K line graph theory, and the opening price and closing price of K line graph are used. The highest price and the lowest price process the shape of K line, then extract the symptom sequence from the fuzzy K chart, then classify the symptom sequence by using data mining technology, and get the statistical features of the fuzzy K line inversion pattern. Finally, the forecast of stock price reversal point is realized by fuzzy K line inversion mode. This paper chooses to conduct empirical research in Shanghai A-share market and Shenzhen A-share market respectively. The results show that the fuzzy K-line inversion model proposed in this paper is real and can be better identified; According to the forecast inversion point of stock price obtained by the fuzzy K-line inversion model proposed in this paper, the return on investment in stock exchange is higher than that in random trading.
【学位授予单位】：湖南大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F830;F224

【相似文献】