Name
Abstract | ||
|---|---|---|
Finding a time series subsequence that is similar to a specific time series is an important problem in trajectory data of vehicles analysis. The problem is made significantly harder for the massive and high-dimensional features of time series. The existing methods for finding the similar subsequences in time series have high time complexity and poor applicability to similar sub-sequence finding of different lengths. In this paper, we propose an acceleration method for similar time-series finding to address this issue. Firstly, our method defines and extracts the feature of the query sequence. Then, we use the feature as the key to search sequence with the same feature to form a candidate set. After that, in each sequence in candidate set, we filter the important points and add it into feature points list to hold the shape characteristics of original sequence better. Finally, Dynamic time warping (DTW) is used to find the similar time-series. Experiment results illustrate that the proposed method can improve the search efficiency and accuracy. |
Year | DOI | Venue |
|---|---|---|
2018 | 10.1007/978-3-030-05081-8_21 | INTERNET OF VEHICLES: TECHNOLOGIES AND SERVICES TOWARDS SMART CITY (IOV 2018) |
Keywords | DocType | Volume |
Time series, Similarity searching, Dimensions reduction, Acceleration method | Conference | 11253 |
ISSN | Citations | PageRank |
0302-9743 | 0 | 0.34 |
References | Authors | |
0 | 5 | |