Scribd
Upload
91 views

Machine Learning Introduction Presentation

This document provides an introduction to machine learning research on time series data. It discusses common machine learning tasks like classification, clustering, and anomaly detection applied to time series. It also covers challenges of time series data like high dimensionality and volume. The document summarizes techniques for speeding up similarity search on time series like dimensionality reduction, indexing, segmentation, and discretization. It discusses alternatives to Euclidean distance for measuring time series similarity, such as cross correlation and dynamic time warping. The document concludes by mentioning other areas of time series research like anomaly detection and change point detection.

Uploaded by

Dino Mandic
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
91 views

Machine Learning Introduction Presentation

This document provides an introduction to machine learning research on time series data. It discusses common machine learning tasks like classification, clustering, and anomaly detection applied to time series. It also covers challenges of time series data like high dimensionality and volume. The document summarizes techniques for speeding up similarity search on time series like dimensionality reduction, indexing, segmentation, and discretization. It discusses alternatives to Euclidean distance for measuring time series similarity, such as cross correlation and dynamic time warping. The document concludes by mentioning other areas of time series research like anomaly detection and change point detection.

Uploaded by

Dino Mandic
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 35

Introduction to Machine Learning Research on Time Series

Umaa Rebbapragada Tufts University Advisor: Carla Brodley 1/29/07

Machine Learning (ML)


Originally a subeld of AI Extraction of rules and patterns from data sets Focused on:
Computational complexity Memory

Machine Learning Tasks for Time Series


Classication Clustering Semi-supervised learning Anomaly Detection

Assumptions
Univariate time series Time series databases

Single Time Series


A single long time series can be converted into a set of smaller time series by sliding a window incrementally across the time series :

Window length is usually a user-specied parameter.

Challenges of Times Series Data


High dimensional Voluminous Requires fast technique

Brute Force Similarity Search


Given query time series Q, the best match by sequential scanning is found by:

O(nd) Finding the nearest neighbor for each time series in the database is prohibitive.

Similarity Search
Clustering and classication methods perform many similarity calculations Some require storage of the k nearest neighbors of each data instance Critical that these calculations be fast

Speeding up Similarity Search


Alternate time series representations Search databases faster New similarity metrics

Data Mining Time Series Toolbox


Indexing Dimensionality Reduction Segmentation Discretization Similarity metric

Indexing
Faster than a sequential scan Insertions and deletions do not require rebuilding the entire index Partition the data into regions Search regions that contain a likely match Requires a similarity metric that obeys triangle inequality

Indexing
R-trees kd-trees linear quad-trees grid-les

Indexing on Times Series Data


High dimensionality slows down speed of computation Curse of dimensionality inhibits efciency of of indexing

Dimensionality Reduction
Reduces the size of the time series Distance on transformed data should lower bound the original distance

This guarantees no false dismissals (false negatives)

Dimensionality Reduction: DFT, DWT, SVD


Represent time series using subsets of
Fourier coefcients Wavelet coefcients eigenvalue/vectors

Euclidean-distance is lower-bounded on DFT1, DWT2, SVD3


[1] C. Faloutsos et al.: Fast Subsequence Matching in Time-Series Databases. SIGMOD Conference 1994: 419-429 [2] K. Chan and A. Fu: Efcient Time Series Matching by Wavelets. ICDE 1999: 126-133 [3] F. Korn et al.: Efciently Supporting Ad Hoc Queries in Large Datasets of Time Sequences. SIGMOD Conference 1997: 289-300

Gemini Framework
Faloutsos et al., 1994 Map each time series to a lower dimension Store in multi-dimensional indexing structure

C. Faloutsos et al.: Fast Subsequence Matching in Time-Series Databases. SIGMOD Conference 1994: 419-429

Piecewise Aggregate Approximation (PAA)

Eamonn J. Keogh, et al.: Dimensionality Reduction for Fast Similarity Search in Large Time Series Databases. Knowl. Inf. Syst. 3(3): 263-286 (2001) Fig: Eamonn J. Keogh, et al.: HOT SAX: Efciently Finding the Most Unusual Time Series Subsequence. ICDM 2005: 226-233

Segmentation
Represent the time series in smaller, less complex segments.
Piecewise Linear Approximation (PLA) Minimum Bounding Rectangles (MBR)

Piecewise Linear Approximation (PLA)

Minimum-Bounding Rectangles (MBR)

Fig: A. Anagnostopoulos et al: Global distance-based segmentation of trajectories. SIGKDD Conference 2006: 34-43

Discretization
Transforms a real-valued time series into a sequence of characters from a discrete alphabet Dimensionality reduction implicit Allows use of string functions on time series

SAX

Jessica Lin et al. A symbolic representation of time series, with implications for streaming algorithms. DMKD 2003: 2-11 Fig: Eamonn J. Keogh, et al.: HOT SAX: Efciently Finding the Most Unusual Time Series Subsequence. ICDM 2005: 226-233

Is Euclidean Distance Best Metric?


Everything discussed so far used ED as similarity metric Is it the best similarity metric for time series?

Drawbacks of Euclidean Distance


Requires two time series to have same dimensionality 1-to-1 alignment of the time axis

Cross Correlation
Cross correlation with convolution can nd optimal phase shift to maximize similarity

Fig: P. Protopapas et al.: Finding outlier light-curves in catalogs of periodic variable stars. Mon. Not. Roy. Astron. Soc. 369 (2006) 677-696

Cross Correlation
Optimal phase shift (to left) of solid line is 0.3

Fig: P. Protopapas et al.: Finding outlier light-curves in catalogs of periodic variable stars. Mon. Not. Roy. Astron. Soc. 369 (2006) 677-696

Dynamic Time Warping (DTW)


DTW allows many-to-one alignment Time series need not be same size

Warped

Time Axis

Fig: Y. Sakurai, et al.: FTW: fast similarity search under the time warping distance. PODS 2005: 326-337 D. J. Berndt, and J. Clifford: Finding Patterns in Time Series: A Dynamic Programming Approach. Advances in Knowledge Discovery and Data Mining 1996: 229-248

DTW Algorithm

DTW Algorithm

Fig: Y. Sakurai, et al.: FTW: fast similarity search under the time warping distance. PODS 2005: 326-337

Drawbacks of DTW
Computationally expensive Does not adhere to triangle inequality => cannot use it for indexing

Making DTW Faster


Global constraints:

Sakoe-Chiba Band

Itakura Parallelogram

Y. Sakurai, et al.: FTW: fast similarity search under the time warping distance. PODS 2005: 326-337

Making DTW Faster


Y. Sakurai et al.: FTW: fast similarity search under the time warping distance. PODS 2005: 326-337 E. Keogh and C. Ratanamahatana: Exact indexing of dynamic time warping. Knowl. Inf. Syst. 7(3): 358-386 (2005) Y. Zhu and D. Shasha: Warping Indexes with Envelope Transforms for Query by Humming. SIGMOD Conference 2003: 181-192 E. Keogh and M. Pazzani: Scaling up dynamic time warping for datamining applications. KDD 2000: 285-289 B.-K. Yi et al.: Efcient Retrieval of Similar Time Sequences Under Time Warping. ICDE 1998: 201-208

Other Areas of Research


Anomaly Detection Change Point Detection

Thesis Research
Anomaly detection methods
fast preserve interesting features

Thank You

You might also like