ASA Competition: Statistical Computing Section
Session Slot: 10:30-12:20 Wednesday
Estimated Audience Size: 100
AudioVisual Request: None
Session Title: Smoothing Methods and Data Analysis
Theme Session: No
Applied Session: Yes
Session Organizer: O'Connell, Michael Becton Dickinson
Address: Becton Dickinson Research Center Research Triangle Park, NC 27709-201
Phone: 919-990-2136
Fax:
Email: Michael_A_OConnell@bdits.bd.com
Session Timing: 110 minutes total (Sorry about format):
TBA (110 minutes total) Opening Remarks by Chair - 5 or 0 minutes First Speaker - 30 minutes (or 25) Second Speaker - 30 minutes Third Speaker - 30 minutes Discussant - 10 minutes (or none) Floor Discusion - 10 minutes (or 5 or 15)
Session Chair: O'Connell, Michael Becton Dickinson
Address: Becton Dickinson Research Center Research Triangle Park, NC 27709-201
Phone: 919-990-2136
Fax:
Email: Michael_A_OConnell@bdits.bd.com
1. When is a Feature Really There? The SiZer Approach
Marron, J.S., University of North Carolina
Address: Department of Statistics University of North Carolina Chapel Hill, NC 27599-3260
Phone: 919-962-2188
Fax: 919-962-1279
Email: marron@stat.unc.edu
Chaudhuri, Probal, Indian Statistical Institute
Abstract: When applying smoothing methods to data, the central question is often ``which observed features are part of the true underlying curve, and which are spurious sampling artifacts?'' An answer is provided using a scale space viewpoint, where one considers the entire family of smooths indexed by the bandwidth. Different bandwidths represent different ``levels of resolution of the data''. Significance of features that are characterized by Zero crossings of the derivative (such as bumps) is easily understood, assessed and presented, simultaneously over all resolutions, via a color map in scale space.
2. FUNFITS: Fitting Functions to Data
Nychka, Douglas, National Center for Atmospheric Research and North Carolina State University
Address: Geophysical Statistics Project National Center for Atmospheric Research Climate and Global Dynamics Division Boulder, CO 80307
Phone: 303-497-1711
Fax: 303-497-1333
Email: nychka@ucar.edu
Abstract: FUNFITS is a suite of functions that enhance the S-PLUS statistical package and facilitate curve and surface fitting. This software grew up as a project to augment S-PLUS with additional nonparametric fitting strategies. Part of this effort was to focus on three areas rich in applications of function estimation and inference: spatial statistics, response surface methodology and nonlinear time series/dynamical systems. This activity has also lead to methods for space filling designs, appropriate when the response is expected to be a complex surface. One unique feature of FUNFITS is the integration of thin plate spline methodology with the more general estimators based on spatial processes. This functionality is implemented completely in the S language and so the algorithms and source code are easily accessible. The core methods implemented in FUNFITS are object driven, including prediction, plot and summary functions that help the user interpret results and visualize the estimate.
3. Some Applications of Spline Smoothing
Gu, Chong, National Institute of Statistical Sciences
Address: National Institute of Statistical Sciences PO Box 14006 Research Triangle Park, NC 27709-4006
Phone: 919-685-9323
Fax: 919-685-9310
Email: gu@hp6.niss.org
Abstract: Nonparametric function estimation procedures are important tools in modern data analysis, of which spline smoothing is among the more effective ones available, especially for multivariate problems. In this talk, we will present some recent developments of spline smoothing in regression, density estimation, and conditional density estimation, and illustrate their applications in data analysis.
List of speakers who are nonmembers: None