Detecting Time Series Motifs Under

1

• Detecting Time Series Motifs Under
• Uniform Scaling
• D. Yankov, E. Keogh, J. Medina, B. Chiu, V.
  Zordan
• Dept. of Computer Science Eng.
• University of California Riverside

2
Outline

• Problem definition
• Motivation
• Formalization and approach
• Experimental evaluation

3
Problem definition

• Given is a long time series or a data set of
  shorter sequences
• Goal
• Detect similar
• patterns of
• various scaling

4
Motivation

• Object recognition with time series
  representation
• Animation

5
Motivation (cont)

• Time series sampled at
• different rate
• Physiological time series of different frequencies

6
Formalization

• Similarity under uniform scaling
• Motifs under uniform scaling

7
Approach

• Observation only a limited set of scaling
  factors need to be checked
• Algorithm.
• For every scaling factor do
• rescale all query subsequences
• represent all time series as equal length words
  over the same alphabet (apply SAX)

8
Approach (cont)

• Using PROJECTION (a locality sensitive hashing
  approach), filter out all non-matching words.
• Compute the distance between the unfiltered time
  series pairs.

9
Experimental evaluation

• Brain activity time series

Valuable in predicting epileptic seizure periods.
10
Experimental evaluation

• Effectiveness of the algorithm
• Efficiency

11
Experimental evaluation (cont)

• Projectile shapes

The algorithm detects a rare cornertang segment
an object that has long intrigued
anthropologists.
12
Experimental evaluation (cont)

• Motion-capture motifs

On this sequence the method detects the same
blocking movement performed by the actor. The
Euclidean distance fails to detect this motif.
13
Conclusion

• Uniform scaling motifs appear in diverse areas as
  animation, object recognition, medical sequence
  mining, etc.
• The presented probabilistic approach for mining
  such motifs is accurate and extremely effective.
• The method works in an entirely unsupervised way,
  requiring only a specified motif length.
• Possible extensions multivariate time series,
  disk resident modifications.

14

• Poster 28
• THANK YOU!