Presentation is loading. Please wait.

Presentation is loading. Please wait.

Abdullah Mueen Eamonn Keogh University of California, Riverside.

Published byAdelia Nash Modified over 8 years ago

Similar presentations

Presentation on theme: "Abdullah Mueen Eamonn Keogh University of California, Riverside."— Presentation transcript:

1 Abdullah Mueen Eamonn Keogh University of California, Riverside

2 Repeated pattern in a time series Utility: – Activity/Event discovery – Summarization – Classification Application – Data Center Chiller Management (Patnaik, et al. KDD 09) – Designing Smart Cane for Elders (Vahdatpour, et al. IJCAI 09) 0 2000400060008000 0 10 20 30

3 Streaming time series Sliding window of the recent history – What minute long trace repeated in the last hour?

4 Discovery 1002003004005006007008009001000 -8000 -7500 -7000... 1 2 3 4 5 6 7 8. 873 The most similar pair of non-overlapping subsequences Maintenance 1 2 3 4 5 6 7 8. 873 874 time:1001 1 2 3 4 5 6 7 8. 873 874 875 time:1002 1 2 3 4 5 6 7 8. 873 874 875 876 time:1003 1 2 3 4 5 6 7 8. 873 874 875 876 877 time:1004 1 2 3 4 5 6 7 8. 873 874 875 876 877 878 time:1005 Update motif pair after every time tick time

5 A subsequence is a high dimensional point – The dynamic closest pair of points problem Closest pair may change upon every update Naïve approach: Do quadratic comparisons. 4 5 2 7 3 6 8 9 4 5 2 7 3 6 8 1 4 5 2 7 3 6 8 1 Deletion Insertion

6 Goal: Algorithm with Linear update time Previous method for dynamic closest pair (Eppstein,00) – A matrix of all-pair distances is maintained O( w 2 ) space required – Quad-tree is used to update the matrix Maintain a set of neighbors and reverse neighbors for all points We do it in O( ) space

7 Methods Evaluation Extensions Case Studies Conclusion

8 Smallest nearest neighbor → Closest pair Upon insertion – Find the nearest neighbor; Needs O( w ) comparisons. Upon deletion – Find the next NN of all the reverse NN 4 5 2 7 3 6 8 1 4 5 2 7 3 6 8 1 Insertion 9 4 5 2 7 3 6 8 1 Deletion 9

9 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 FIFO Sliding window N-lists Neighbors in order of distances 8 8 7 7 2 2 1 1 8 8 1 1 2 2 4 4 5 5 1 1 1 1 7 7 3 3 3 3 6 6 7 7 7 7 6 6 7 7 5 5 2 2 4 4 5 5 5 5 1 1 4 4 8 8 3 3 7 7 5 5 7 7 2 2 4 4 8 8 4 4 2 2 5 5 2 2 1 1 3 3 6 6 1 1 3 3 8 8 1 1 8 8 3 3 8 8 3 3 5 5 6 6 6 6 4 4 6 6 4 4 6 6 R-lists Points that should be updated Total number of nodes in R-lists is w 5 5 1 1 3 3 2 2 4 4 8 8 6 6 7 7 4 5 2 7 3 6 8 1 4 4 7 7 Still O( w 2 ) space (ptr, distance)

10 While inserting  Updating NN of old points is not necessary  A point can be removed from the neighbor list if it violates the temporal order Average length O( )

11 4 5 2 7 3 6 8 1 1 1 1 1 2 2 1 1 3 3 1 1 2 2 5 5 2 2 8 8 3 3 2 2 4 4 3 3 5 5 3 3 6 6 4 4 7 7 5 5 6 6 4 4 7 7 6 6 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 2 2 3 3 3 3 2 2 3 3 2 2 6 6 3 3 7 7 9 9 5 5 5 5 4 4 4 4 6 6 8 8 5 5 7 7 6 6 6 6 8 8 4 4 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 1 1 4 5 2 7 3 6 8 1 9 4 5 2 7 3 6 8 9 10 3 3 3 3 4 4 3 3 6 6 6 6 5 5 7 7 9 9 5 5 4 4 4 4 7 7 8 8 5 5 6 6 4 4 6 6 8 8 5 5 7 7 8 8 9 9 3 3 4 4 5 5 6 6 7 7 8 8 9 9 2 2 1 1

12 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 Fast Pair Insect EEG EOG RW Average Update Time (Sec) 00.511.522.533.54 x 10 4 Window Size (w) 02004006008001000 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Insect EEG EOG RW Fast Pair Average Update Time (Sec) Motif Length (m) 00.511.522.533.54 x 10 4 10 20 30 40 50 60 70 80 90 100 110 Insect EEG EOG RW Average Length of N-list Window Size (w) 02004006008001000 0 50 100 150 200 250 300 350 Insect EEG EOG RW Motif Length (m) Average Length of N-list

13 Methods Evaluation Extensions Case Studies Conclusion

14 Multidimensional Motif Maintain motif in Multiple Synchronous time series Points in one series can have neighbors in others Arbitrary Data Rate Load shedding: Skip points if can’t handle 0.40.50.60.70.80.91.0 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Fraction of Data Used Fraction of Motifs Discovered EOG Random Walk EEG Insect FIFO for each streams

15 Replace all the occurrences of a motif by a pointer to that motif. For signals with regular repetitions – Higher compression rate with less error. Dictionary

16 Check if a robot closed a loop Convert the stream of video frames to stream of color histograms. Most similar color histograms are good candidates for loop detection. 040080012001600 0 4 8 12 16 0 40080012001600 0 4 8 12 16 Start Loop Closure Detected

17 First attempt to maintain time series motif online – Maintains minutes long repetition in hours long sliding window Linear update time with less space cost than quad-tree based method – O( ) Vs O(w 2 ) Faster than general dynamic closest pair solution

18 Code and Data: http://www.cs.ucr.edu/~mueen/OnlineMotif

Download ppt "Abdullah Mueen Eamonn Keogh University of California, Riverside."

Similar presentations

Indexing Time Series Based on original slides by Prof. Dimitrios Gunopulos and Prof. Christos Faloutsos with some slides from tutorials by Prof. Eamonn.

Indexing Time Series Based on original slides by Prof. Dimitrios Gunopulos and Prof. Christos Faloutsos with some slides from tutorials by Prof. Eamonn.
SAX: a Novel Symbolic Representation of Time Series

SAX: a Novel Symbolic Representation of Time Series
Probabilistic Skyline Operator over Sliding Windows Wenjie Zhang University of New South Wales & NICTA, Australia Joint work: Xuemin Lin, Ying Zhang, Wei.

Probabilistic Skyline Operator over Sliding Windows Wenjie Zhang University of New South Wales & NICTA, Australia Joint work: Xuemin Lin, Ying Zhang, Wei.
Image classification Given the bag-of-features representations of images from different classes, how do we learn a model for distinguishing them?

Image classification Given the bag-of-features representations of images from different classes, how do we learn a model for distinguishing them?
Incremental Clustering Previous clustering algorithms worked in “batch” mode: processed all points at essentially the same time. Some IR applications cluster.

Incremental Clustering Previous clustering algorithms worked in “batch” mode: processed all points at essentially the same time. Some IR applications cluster.
Doruk Sart, Abdullah Mueen, Walid Najjar, Eamonn Keogh, Vit Niennatrakul 1.

Doruk Sart, Abdullah Mueen, Walid Najjar, Eamonn Keogh, Vit Niennatrakul 1.
Searching on Multi-Dimensional Data

Searching on Multi-Dimensional Data
Distributed Indexed Outlier Detection Algorithm Status Update as of March 11, 2014.

Distributed Indexed Outlier Detection Algorithm Status Update as of March 11, 2014.
Efficient Anomaly Monitoring over Moving Object Trajectory Streams joint work with Lei Chen (HKUST) Ada Wai-Chee Fu (CUHK) Dawei Liu (CUHK) Yingyi Bu (Microsoft)

Efficient Anomaly Monitoring over Moving Object Trajectory Streams joint work with Lei Chen (HKUST) Ada Wai-Chee Fu (CUHK) Dawei Liu (CUHK) Yingyi Bu (Microsoft)
1 NNH: Improving Performance of Nearest- Neighbor Searches Using Histograms Liang Jin (UC Irvine) Nick Koudas (AT&T Labs Research) Chen Li (UC Irvine)

1 NNH: Improving Performance of Nearest- Neighbor Searches Using Histograms Liang Jin (UC Irvine) Nick Koudas (AT&T Labs Research) Chen Li (UC Irvine)
Mining Time Series.

Mining Time Series.
Introduction to Bioinformatics

Introduction to Bioinformatics
Common approach 1. Define space: assign random ID (160-bit) to each node and key 2. Define a metric topology in this space,  that is, the space of keys.

Common approach 1. Define space: assign random ID (160-bit) to each node and key 2. Define a metric topology in this space,  that is, the space of keys.
VLSH: Voronoi-based Locality Sensitive Hashing Sung-eui Yoon Authors: Lin Loi, Jae-Pil Heo, Junghwan Lee, and Sung-Eui Yoon KAIST

VLSH: Voronoi-based Locality Sensitive Hashing Sung-eui Yoon Authors: Lin Loi, Jae-Pil Heo, Junghwan Lee, and Sung-Eui Yoon KAIST
A Novel Scheme for Video Similarity Detection Chu-Hong Hoi, Steven March 5, 2003.

A Novel Scheme for Video Similarity Detection Chu-Hong Hoi, Steven March 5, 2003.
Automatic Identification of ROIs (Regions of interest) in fMRI data.

Automatic Identification of ROIs (Regions of interest) in fMRI data.
Stabbing the Sky: Efficient Skyline Computation over Sliding Windows COMP9314 Lecture Notes.

Stabbing the Sky: Efficient Skyline Computation over Sliding Windows COMP9314 Lecture Notes.
© University of Minnesota Data Mining for the Discovery of Ocean Climate Indices 1 CSci 8980: Data Mining (Fall 2002) Vipin Kumar Army High Performance.

© University of Minnesota Data Mining for the Discovery of Ocean Climate Indices 1 CSci 8980: Data Mining (Fall 2002) Vipin Kumar Army High Performance.
Themis Palpanas1 VLDB - Aug 2004 Fair Use Agreement This agreement covers the use of all slides on this CD-Rom, please read carefully. You may freely use.

Themis Palpanas1 VLDB - Aug 2004 Fair Use Agreement This agreement covers the use of all slides on this CD-Rom, please read carefully. You may freely use.
Abdullah Mueen UC Riverside Suman Nath Microsoft Research Jie Liu Microsoft Research.

Abdullah Mueen UC Riverside Suman Nath Microsoft Research Jie Liu Microsoft Research.

Similar presentations

Ads by Google