|
Activity Number:
|
73
- Nonparametric Statistics in High-Dimensional Settings
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 30, 2017 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Nonparametric Statistics
|
|
Abstract #323655
|
View Presentation
|
|
Title:
|
Nonparametric non-convex penalized regression spline and application in ECG recording
|
|
Author(s):
|
Yao Chen* and Xiao Wang and Mohammad Adibuzzaman and Yonghan Jung
|
|
Companies:
|
Purdue University and Purdue University and Purdue University and Purdue University
|
|
Keywords:
|
Regression Spline ;
non-convex penalty ;
KKT condition ;
ECG recording
|
|
Abstract:
|
Regression spline is useful in nonparametric regression. In this paper, a regression spline with non-convex penalty is used, and we propose a new regularization technique to choose optimal knots and smoothing parameter. At the same time, it produces smoothing spline with optimal converge rate. An iterative algorithm derived from KKT condition is used to estimate the unknown coefficient. Simulation and real data analysis of ECG recording (classification) have shown a superior performance of our method against many existing approaches.
|
Authors who are presenting talks have a * after their name.
