|
Activity Number:
|
457
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #303448 |
|
Title:
|
Multiple Curve-Fitting with BARS
|
|
Author(s):
|
Sam Behseta*+ and Robert E. Kass
|
|
Companies:
|
California State University, Fullerton and Carnegie Mellon University
|
|
Address:
|
Mathematics Department, Fullerton, CA, 92834,
|
|
Keywords:
|
BARS ; Functional Data ; Curve-Fitting
|
|
Abstract:
|
We present two methods for fitting curves to a population of histograms. Both methods utilize the properties of Bayesian Adaptive Regression Splines (BARS) to obtain the fits either simultaneously or individually. Some applications associated with multiple curve-fitting will be discussed. First, a hierarchical model is formed to assess the variability between the curves while accounting for the variability due to the individual curve estimation. Second, a method is proposed to test the hypothesis of the equality of two functions. Finally, we consider the problem of comparing several noisy functions. We propose testing procedures to perform comparisons either pointwise or globally over the entire length of functions.
|
