Nonparametric Inference Tutorial
This is the web page to accompany
the nonparametric inference tutorial at the
SAMSI Opening Workshop for the Astrostatistics Program.
The tutorial took place on January 21 and 22, 2006,
led by Larry Wasserman and Chad Schafer, members of the International Computational
Astrostatistics Group.
You can get the slides we used here. We are going to fix
some of the typos we found and repost soon.
Topics include:
- General Concepts of Smoothing
- Bias-Variance Tradeoff
- Linear Smoothers
- Cross Validation
- Local Polynomial Regression, Kernels and Splines
- Confidence Bands
- Multiple Regression
- Density Estimation
- Orthogonal Functions
- Wavelets
- Measurement Error
- Inverse Probems
- Classification
IMPORTANT:Things to do Prior to the Tutorial
The tutorial will include an overview of how to use the R statistical
package to utilize the nonparametric methods. Although not
necessary, participants would find it helpful to bring a laptop
computer set up with the necessary software and files:
- Install R on your laptop. You can find R at
the R project main page.
- Install required packages.
Once you have installed R, start it up, and
choose "Install package(s) from CRAN..." from under the "Packages"
menu. Choose each of the following packages from the list:
- fda
- locfit
- tree
- wavethresh
We will be using these during the tutorial.
- Download
these files to your computer.
R Tutorials
We will not assume you know anything regarding R, but looking
these over would be helpful:
R Tutorial from Center for Astrostatistics at Penn State.
R Tutorial by Phil Spector
Data Sets
The WMAP data.
Some astronomical data sets from the Center for Astrostatistics
at Penn State.
Related Reading
We will fill this in soon with some references.
Astronomy Applications
Nonparametric Inference for the Cosmic Microwave Background
by Genovese, et.al.
Empirical models for Dark Matter Halos
by Graham, et.al.
Contact
Please contact Chad Schafer at cschafer at stat dot cmu dot edu
if you have questions.