Discovering Subdimensional Motifs of Different Lengths in LargeScale Multivariate Time Series
Abstract
Detecting repeating patterns of different lengths in time series, also called variablelength motifs, has received a great amount of attention by researchers and practitioners. Despite the significant progress that has been made in recent single dimensional variablelength motif discovery work, detecting variablelength \textit{subdimensional motifs}patterns that are simultaneously occurring only in a subset of dimensions in multivariate time seriesremains a difficult task. The main challenge is scalability. On the one hand, the bruteforce enumeration solution, which searches for motifs of all possible lengths, is very time consuming even in single dimensional time series. On the other hand, previous work show that indexbased fixedlength approximate motif discovery algorithms such as random projection are not suitable for detecting variablelength motifs due to memory requirement. In this paper, we introduce an approximate variablelength subdimensional motif discovery algorithm called \textbf{C}ollaborative \textbf{HI}erarchy based \textbf{M}otif \textbf{E}numeration (CHIME) to efficiently detect variablelength subdimensional motifs given a minimum motif length in largescale multivariate time series. We show that the memory cost of the approach is significantly smaller than that of random projection. Moreover, the speed of the proposed algorithm is significantly faster than that of the stateoftheart algorithms. We demonstrate that CHIME can efficiently detect meaningful variablelength subdimensional motifs in large real world multivariate time series datasets.
 Publication:

arXiv eprints
 Pub Date:
 November 2019
 arXiv:
 arXiv:1911.09218
 Bibcode:
 2019arXiv191109218G
 Keywords:

 Computer Science  Machine Learning;
 Computer Science  Artificial Intelligence
 EPrint:
 Accepted by ICDM 2019