|
Activity Number:
|
150
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #309039 |
|
Title:
|
Approaches to Time Series Clustering
|
|
Author(s):
|
Hwanseok Choi*+ and J. Michael Hardin
|
|
Companies:
|
The University of Alabama and The University of Alabama
|
|
Address:
|
PO Box 870221, Tuscaloosa, AL, 35487-0221,
|
|
Keywords:
|
Time Series Clustering ; MPCA ; Multivariate Time series ; Dimension Extraction
|
|
Abstract:
|
Time series clustering has been one of the major techniques in providing useful information in various fields these days especially as applied to data mining. There seems to be an increased interest in time series clustering in various domains. However, the majority of time series clustering studies are restricted to univariate time series; and, there is lack of studies to compare different time series clustering approaches. In this presentation we introduce several feature extraction approach models to reduce original time series dimensions. These techniques can be used to draw lower dimension clustering plots with multivariate time series data. After that, we compare the results of predictability among multivariate time series clustering methods with simulation data as well as field data. Further, we examine these cluster methodologies across different time series contexts.
|
