Activity Number:
|
145
- Trend Filtering and Related Regression Methods
|
Type:
|
Invited
|
Date/Time:
|
Tuesday, August 10, 2021 : 10:00 AM to 11:50 AM
|
Sponsor:
|
IMS
|
Abstract #316781
|
|
Title:
|
Dyadic CART Revisited
|
Author(s):
|
Sabyasachi Chatterjee*
|
Companies:
|
University of Illinois at Urbana-Champaign
|
Keywords:
|
|
Abstract:
|
Proposed by Donoho (1997), Dyadic CART is a nonparametric regression method which computes a globally optimal dyadic deci- sion tree and fits piecewise constant functions in two dimensions. We revisit Dyadic CART and related methods in the context of estimation of multivariate piecewise polynomial signals. We show these methods are minimax rate optimal and computationally efficient in the fixed design set up. We also study these estimators in the context of Trend Filtering and make comparisons.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2021 program
|