Comparison of Subsequence Pattern Matching Methods for Financial Time Series

In contrast to general time series analysis, only a few numbers of studies are devoted to subsequence pattern matching methods for financial time series. In this paper, we compare the processing time and accuracy of three well-known pattern matching methods from financial time series domain and two pattern matching methods from general time series area. Our experiment was conducted on the historical data of Hang Seng Index (HSI) from Hong Kong Stock Market. Our experiment reveals that segmentation step and time distortion issues can significantly affect the performance of these methods.