A new implementation of similar segment search in an arbitrary number of time-series
暂无分享,去创建一个
This paper proposes a new implementation of a similar segment search algorithm in an arbitrary number of time-series. Recursive diamond division search (RDDS) can theoretically search for similar segments in any arbitrary number of time-series. Previous implementations were based on the number of loops, which corresponded to the number of the given time-series; RDDS-2 and RDDS-3 were implemented for two time-series and three time-series, respectively. This new implementation RDDS-n can process an arbitrary number of time-series. For three time-series, RDDS-n with n = 3 achieves the same processing time as RDDS-3. The new implementation is evaluated with 3, 4 and 5 time-series to show its effectiveness.
[1] Sugiyama Masahide,et al. An Efficient Algorithm for Similar Segment Search -- Implementation and Evaluation in Three Time-Series , 2007 .
[2] Kunio Kashino,et al. Quick audio retrieval based on histogram feature sequences. , 2000 .
[3] Masahide Sugiyama. A New Similar Segment Search Algorithm in Multiple Time Series , 2006, 2006 IEEE Workshop on Multimedia Signal Processing.
[4] Sugiyama Masahide. On Diamond Covering in Similar Segment Search RDDS Algorithm , 2007 .