|
Activity Number:
|
321
- Detecting Structural Change in Complex Data
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, July 31, 2018 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
International Chinese Statistical Association
|
|
Abstract #326565
|
Presentation
|
|
Title:
|
Penalized Versus Segmentation Methods in Changepoint Problems
|
|
Author(s):
|
Ryan Tibshirani* and Sangwon Hyun and Kevin Lin and Max G'Sell
|
|
Companies:
|
Carnegie Mellon University and Carnegie Mellon University and Carnegie Mellon University and Carnegie Mellon University
|
|
Keywords:
|
Fused lasso;
Binary segmentation;
Post-selection inference;
Selective inference
|
|
Abstract:
|
Roughly speaking, penalization methods and segmentation methods represent two major classes of estimators in changepoint problems. This talk will briefly cover some of the benefits and drawbacks of each class. We will focus on the simple and classical 1-dimensional mean changepoint problem, but will also cover more general (more challenging) changepoint problem settings. Time permitting, we will discuss post-selection inference with these estimators.
|
Authors who are presenting talks have a * after their name.
