Discover motifs in multi-dimensional time-series using the principal component analysis and the MDL principle

Authors:
Yoshiki Tanaka;Kuniaki Uehara
Affiliations:
Department of Computer and Systems Engineering, Kobe University, Kobe, Japan;Department of Computer and Systems Engineering, Kobe University, Kobe, Japan
Venue:
MLDM'03 Proceedings of the 3rd international conference on Machine learning and data mining in pattern recognition
Year:
2003

Citing 5
Cited 18

Stochastic Complexity in Statistical Inquiry Theory

Stochastic Complexity in Statistical Inquiry Theory
Distance Measures for Effective Clustering of ARIMA Time-Series

ICDM '01 Proceedings of the 2001 IEEE International Conference on Data Mining
On the Discovery of Weak Periodicities in Large Time Series

PKDD '02 Proceedings of the 6th European Conference on Principles of Data Mining and Knowledge Discovery
Principal Component Analysis on Vector Computers

VECPAR '96 Selected papers from the Second International Conference on Vector and Parallel Processing
Extraction of Primitive Motion and Discovery of Association Rules from Human Motion Data

Progress in Discovery Science, Final Report of the Japanese Discovery Science Project

Visually mining and monitoring massive time series

Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining
Discovery of Time-Series Motif from Multi-Dimensional Data Based on MDL Principle

Machine Learning
Visualizing and discovering non-trivial patterns in large time series databases

Information Visualization
Recovery of missing information in graph sequences by means of reference pattern matching and decision tree learning

Pattern Recognition
Experiencing SAX: a novel symbolic representation of time series

Data Mining and Knowledge Discovery
Numerical time-series pattern extraction based on irregular piecewise aggregate approximation and gradient specification

New Generation Computing
An efficient stream mining technique

WSEAS Transactions on Information Science and Applications
Time series classification based on qualitative space fragmentation

Advanced Engineering Informatics
An efficient time series data mining technique

ICCOMP'08 Proceedings of the 12th WSEAS international conference on Computers
On-line motif detection in time series with SwiftMotif

Pattern Recognition
Motif detection inspired by immune memory

ICARIS'07 Proceedings of the 6th international conference on Artificial immune systems
Detecting motifs in system call sequences

WISA'07 Proceedings of the 8th international conference on Information security applications
Temporal data mining using shape space representations of time series

Neurocomputing
Discovery of skills from motion data

JSAI'03/JSAI04 Proceedings of the 2003 and 2004 international conference on New frontiers in artificial intelligence
Identifying predictive multi-dimensional time series motifs: an application to severe weather prediction

Data Mining and Knowledge Discovery
Mining approximate motifs in time series

DS'06 Proceedings of the 9th international conference on Discovery Science
Comparison of two different prediction schemes for the analysis of time series of graphs

IbPRIA'05 Proceedings of the Second Iberian conference on Pattern Recognition and Image Analysis - Volume Part II
Analysis of time series of graphs: prediction of node presence by means of decision tree learning

MLDM'05 Proceedings of the 4th international conference on Machine Learning and Data Mining in Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

Recently, the detection of a previously unknown, frequently occurring pattern has been regarded as a difficult problem. We call this pattern as "motif". Many researchers have proposed algorithms for discovering the motif. However, if the optimal period length of the motif is not known in advance, we cannot use these algorithms for discovering the motif. In this paper, we attempt to dynamically determine the optimum period length using the MDL principle. Moreover, in order to apply this algorithm to the multi dimensional time-series, we transform the time-series into one dimensional time-series by using the Principal Component Analysis. Finally, we show experimental results and discuss the efficiency of our motif discovery algorithm.