A Novel Method for Fast and Accurate Similarity Measure in Time Series Field

摘要

Similarity measure is a central problem in time series data mining. Although most approaches to this problem have been developed, with the rapid growth of the amount of data, we believe there is a challenging demand for supporting similarity measure in a fast and accurate way. In this paper, we propose a new time series representation model and a corresponding similarity measure, which is able to capture the main trends of time series and fulfill fast similarity detection. We compare the new method with state-of-the-art time series similarity methods and dimension-reduction techniques to indicate its superiority. Experiment results demonstrate the new method is able to support both fast and accurate similarity measure.