|
Activity Number:
|
412
- Theory and Methods for Change-Point and Abnormality Detection
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2018 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
IMS
|
|
Abstract #330730
|
Presentation
|
|
Title:
|
Change-Point Estimation of Trend in High-Dimensional Time Series
|
|
Author(s):
|
Monika Bhattacharjee * and Moulinath Banerjee and George Michailidis
|
|
Companies:
|
University of Florida and University of Michigan and University of Florida
|
|
Keywords:
|
change-point;
trend;
estimation;
high-dimension;
time series
|
|
Abstract:
|
We consider signal-plus-noise high-dimensional time series model where noise is a mean zero stationary process and signal or trend function belongs to some parametric family. This family includes a wide range of trend functions such as linear, polynomial, log-linear and Gompertz functions. Suppose a change-point occurs due to the change in parameters associated with trend function. We provide a consistent estimator of the change point and establish its asymptotic distribution. We also compare our results with other existing works in the literature which deal with specific trend functions.
|
Authors who are presenting talks have a * after their name.
