|
Statistics 526 Term 2 1998-99
Smoothing Methods in Regression
Handouts/Course Notes
January 5: Birthweight data in-class handout
gives examples of parametric and gam analysis of 0-1 data, to estimate the probability
of survival of newborns as a function of birthweight and gestation period.
For data:
Description,
Data
January 12: Methods in-class handout
describes some smoothing methods and applies them to a fictitious data
set.
After lecture January 12: Nadaraya-Watson Estimate Part 1
definition, examples, asymptotics for a simple case. Not distributed in lecture.
January 19: Nadaraya-Watson Estimate
contains January 12 Nadaray-Watson blurb (with minor corrections), plus two new sections on general
asymptotics.
Not distributed in lecture.
January 26: Splus function myksmooth, to calculate the Nadaraya-Watson
estimate using a kernel of order 4
February 2:
Number of parameters: Splus work on looking
at the trace of the hat matrix;
Smoothing Parameter: Choosing the
smoothing parameter using the truth, using the data and plug-in
methods
Smoothing Parameter: Write up on choosing the
smoothing parameter (not distributed in class). Cross-validation
section has been added (Feb 9).
February 9:
CV/GCV: in smooth.spline in Splus.
February 23
CV/GCV/plug-in/truth: locpoly simulations in Splus: short
version, complete version (not
distributed).
Summary of local polynomial estimation (distributed),
long version (not distributed).
March 2
Multiple Regression
Kernel Methods;
Generalized Additive Models,
Generalized Additive Models - Splus
March 9
Hatmatrix for 2-D loess, bivariate gam (not distributed)
Parametric Model Fitting tex write-up (not distributed)
Model Fitting in Gam
March 16, 23
General Linear Models and General Additive Models
March 23
Splus examples of model
checking with generalized additive models
March 30
Least squares regression and
Bspline regression.
Smoothing Splines and Penalized Likelihood.
Smoothing spline
examples in Splus.
|
