针对金融时间序列分析中注重快速作出趋势判断的特点,利用数据挖掘的思想和工具,提出

一种金融时间序列模式快速发现算法. 与传统的预测算法相比较,该算法对数据的分布和平稳性等

方面的要求不高,不基于任何假设,能够非常快速地发现时间序列中的频繁模式,经过模式匹配后,

可以用于金融时间序列的分析与预测. 以实际汇率数据为例,证明了该算法的有效性-Financial time series analysis for rapid prediction of trend-oriented features, use of data mining ideas and tools of financial time series model is proposed Fast algorithm for discovering. With the traditional prediction method compared to the algorithm of data distribution and stability etc. do not ask, do not based on any assumptions, can very quickly find that the frequent time series patterns, by matching patterns can be used for financial time series analysis and prediction. to exchange rate data, show that the algorithm