Time Series Similar Pattern Matching Based on Empirical Mode Decomposition

Similar pattern matching of sequence is an important field in time series data mining. Since time series may be very long, which results in query performance decreasing sharply when the database is large, therefore, dimension reduction is required before pattern matching. Fourier transform can be used for dimension reduction, but it cannot provide any feature of signals in local interval. According to this situation, a new similar pattern matching method is proposed in this paper. Firstly, trends of time series are extracted by empirical mode decomposition, and the trends are translated into vectors to realize dimension reduction. Secondly, the vectors are clustered by a forward propagation learning algorithm. Finally, all the series that are similar with the query are found by calculating Euclidean distance between the query and the series that belong to the same category with it. Experimental results show that it is an effective pattern-matching algorithm