Welcome to CSI 979

Generalized Additive Models

Instructor: James Gentle

Spring, 1996

The class meets on Wednesdays at 4:30 in Science & Technology 126.

PostScript versions of the lectures, ASCII files of data and S-Plus functions, and other files are available.

We will use the book Generalized Additive Models, by Hastie and Tibshirani, and chapters from Computational Statistics, by Gentle, which will be made available electronically (in /Users/jgentle/book on science).

Student work in the course (and the relative weighting of this work in the overall grade) will consist of

  • frequent, small assignments (computing problems, etc.) (15)
  • a semester project (30)
  • an in-class midterm (25)
  • a final exam consisting of an in-class component and a take-home component (30)
    • The evaluation of an assignment will be done in the standard way:
    Is it on time and does it meet specs?

    Lectures and Assignments

    (The assignments will be posted as the semester proceeds.)

    January 24, 1996

    Lecture 1 (PostScript) Introduction to generalized additive models; Monte Carlo methods; S-Plus introduction; smoothing

    January 31, 1996

    Lecture 2 (PostScript) Smoothing
  • Project milestone: report on two articles that use Monte Carlo studies in generalized additive models.

    February 7, 1996

    Lecture 3 (PostScript) Basics of generalized linear models, link functions
  • Project milestone: Design a plan to replicate and extend one of the studies

    February 14, 1996

    Presentations and discussions of plans for projects.
  • Project milestone: Feasibility study -- software, etc.

    February 21, 1996

    Lecture 4 (PostScript) Generalized linear models and binary data; fitting models by ML
    Assignment

    February 28, 1996

    Lecture 5 (PostScript) Generalized linear models and binary data; an example
    Assignment
    cancer.dat

    March 6, 1996

    Lecture 6 (PostScript) Residuals and other topics

    March 13, 1996

    GMU recess

    March 20, 1996

  • Project milestone: Web page preliminary report
    Lecture 7 (PostScript) Modeling: Various topics.

    March 27, 1996

    Midterm (PostScript)

    April 3, 1996

    Catchup: Review; Review of midterm (PostScript)
  • Project milestone: Review and critique a fellow student's paper

    April 10, 1996

    Lecture 9: Response transformations

    April 17, 1996

    Lecture 10 (PostScript) Response transformations
  • Project milestone: Complete the revision of paper based on review

    April 24, 1996

    Lecture 11 (PostScript) Topics in applications of generalized additive models
  • Project milestone: Final project Web page up

    May 1, 1996

  • Project milestone: Presentations
    Handout take-home portion of final exam
    Data

    May 8, 1996

    Exam (in-class)

    Computational Resources

    Labs with Unix workstations are available for use in this class in both CSI and SITE.
  • CSI facilities.
  • Software available in SITE labs.

    It will be necessary to use Unix, C or Fortran, and various application packages in this course. The Intel Paragon computer will also be used. Some notes on these systems are available.

    General Reference Materials

  • Some general -- but incomplete! -- stuff on software.
  • GAMS categorized list of software
  • GAMS, general (Guide to Available Software, NIST)
  • Unix
  • Fortran 90
  • C
  • Emacs
  • html
  • html conversion filters
  • GhostScript
  • GhostScript (NCSA)
  • PostScript
  • Matlab (courtesy of Indiana University Center for Statistical and Mathematical Computing)
    The GMU Copy Shop also has an inexpensive Matlab primer.
  • S (or S-Plus)
  • IMSL Fortran Libraries
  • Maple (courtesy of Indiana University Center for Statistical and Mathematical Computing)

    Reference Material for Local Access

  • science.gmu.edu info
  • Paragon local info
  • emacs local info
  • TeX, LaTeX local info
  • latex2html local info
  • PostScript local info
  • Fortran 90 local info
  • IMSL Fortran Libraries info
  • IMSL C Libraries info
  • Matlab local info
  • PV-Wave local info
  • Maple local info
  • S-Plus local info

    Other Resources

    The most important WWW repository of statistical stuff (datasets, programs, general information, connection to other sites, etc.) is StatLib Index at Carnegie Mellon.

    Student Home Pages

    The students in the class all have homepages on which they put parts of their assignments and other interesting stuff.
  • Amrut Champaneri
  • Michael B. Feil
  • Arndt Laemmerzahl
  • Jim Shine

    James Gentle, jgentle@gmu.edu