Activity Number:
|
175
- Clustering and Changepoint Analysis
|
Type:
|
Contributed
|
Date/Time:
|
Monday, July 29, 2019 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Korean International Statistical Society
|
Abstract #304253
|
Presentation
|
Title:
|
High-Dimensional Changepoint Detection via a Geometrically Inspired Mapping
|
Author(s):
|
Thomas Grundy* and Rebecca Killick and Gueorgui Mihaylov
|
Companies:
|
STOR-i Centre for Doctoral Training, Lancaster University and Lancaster University, UK and Royal Mail/GBI Data Science Group
|
Keywords:
|
changepoint;
high-dimensional;
multivariate;
time series
|
Abstract:
|
High-dimensional changepoint analysis is a growing area of research and has applications in a wide range of fields. The aim is to accurately and efficiently detect changepoints in time series data as the number of time points and dimensions grows large. Existing methods typically aggregate or project the data to a smaller number of dimensions, usually one, before performing a univariate analysis. We present a high-dimensional changepoint detection method that takes inspiration from geometry to map the high-dimensional time series to two dimensions. Applying univariate changepoint detection methods to both dimensions allows the detection of changepoints that correspond to changes in the mean and variance of the original time series. We demonstrate that this approach outperforms the current state-of-the-art high-dimensional changepoint methods both in the accuracy of detected changepoints and computational efficiency.
|
Authors who are presenting talks have a * after their name.