[S] S-PLUS User Conference and One Day Workshops

Lisa Eaton (lisae@statsci.com)
Wed, 01 Jul 1998 11:47:23 -0700


1998 S-PLUS User Conference One Day Workshops
Washington DC

MathSoft invites you to participate in the 1998 International S-PLUS User
Conference in Washington, DC, October 8-9, 1998. The conference is a forum
for S-PLUS users from all industries to exchange ideas on data analysis,
visualization and modeling with S-PLUS. Conference presentations will
feature specific user applications and data analysis techniques for a wide
variety of industries, including pharmaceuticals, finance, semiconductors
and others.

In addition to the conference program, MathSoft is also offering a series
of one-day workshops. Users have the opportunity to learn cutting-edge
programming and data analysis techniques from well-known S-PLUS pioneers.

For more information about the S-PLUS User Conference and the one-day
workshops, please go to the MathSoft website at:
www.mathsoft.com/splus/splsprod/userconf.htm.

"Efficient Programming in S"
Presenter: Dr. Bill Venables, University of Adelaide
Co-presenter: Prof. Brian Ripley, University of Oxford
October 6, 1998
9am - 4pm

The central theme of this workshop will be an introduction and overview to
the new language developments in S version 4, which is to become the 'S
engine' driving forthcoming versions of S-PLUS. Although
back-compatibility should be nearly complete, there are many valuable new
features to the language that require some changes to programming style in
order to exploit fully.

The workshop will also cover standard, if advanced, programming techniques
that will apply to old and new engines, such as writing model fitting
functions, manipulating the language itself, advanced object orientation.
It will also cover some practical contemporary programming topics,
programming for the GUI in S-PLUS 4.5 and issues in working with Unix and
Windows such as using compiled code, packaging software and producing help
information.

"Modern Data Analysis in S-PLUS"
Presenter: Prof. Brian Ripley, University of Oxford
Co-presenter: Dr. Bill Venables, University of Adelaide
October 7, 1998
9am - 4pm

The aim of this workshop is to explore the richness of modern statistical
techniques that are available either in S-PLUS or through contributed
packages, as well as some of the issues in developing such packages. The
primary audience is S-PLUS users who would like to see how it can be used
in modern statistics, especially those who would like to package their own
statistical tools for a wider audience.

Topics to be covered (with case-studies) will draw from visualization
(density estimation, smoothing, multi-dimensional scaling), classification
and survival trees (using RPART), "beyond linear regression" (additive
models, MARS, projection pursuit regression), neural networks, discriminant
analysis and other techniques from pattern recognition.

"Strategies for Developing and Validating Predictive Models"
Presenter: Dr. Frank Harrell, University of Virginia
October 10, 1998
9am - 5pm

Regression models must be flexible enough to fit nonlinear and non-additive
relationships, but unless the sample size is enormous, the approach to
modeling must avoid data dredging (mining), which results in overfitting
and a failure of the predictive model to validate on new patients.

The first half of the course presents the following elements of
multivariable predictive modeling: using regression splines to relax
linearity assumptions, perils of variable selection and overfitting,
shrinkage, missing data, data reduction, interaction surfaces, and model
validation by re-sampling. The afternoon half of the course will put the
modeling strategy together, demonstrating methods for:
(1) understanding relationships among predictor variables,
(2) uncovering patterns of missing data,
(3) displaying effects of predictors,
(4) presenting models graphically to non-statisticians, and
(5) using the bootstrap to internally validate the model while accounting
for overfitting.

"Financial Topics in S-PLUS"
Presenter: Dr. Doug Martin, MathSoft, Inc.
October 10, 1998
9am - 4pm

This course is designed to help you effectively utilize the S-PLUS
financial analysis and modeling software to develop your own cutting edge
financial models. S-PLUS is the most sophisticated software environment
available for this purpose today, by virtue of:
- its modern object-oriented programming base for supporting very rapid
prototyping and development of proprietary trading models
- a very large collection of statistical modeling and visualization tools
- exceptionally modern methods flowing from recent fundamental research
results, e.g., the new modern robust regression method in S-PLUS 4.5.

Topics in this course include:
- Non-Gaussian financial returns
- Building univariate or multivariate GARCH volatility model(s) to
accurately measure your instrument and portfolio value-at-risk (VaR)
- New robust regression methods to calculate robust estimates of beta
- Recently developed influence functions and robust covariance matrix
estimates for your portfolio construction and sensitivity analysis

1998 International S-PLUS User Conference, October 8-9, 1998 in Washington, DC

Including One Day Workshops:
"Effective Programming in S" with Bill Venables and Brian Ripley
"Modern Data Analysis Techniques in S-PLUS" with Brian Ripley and Bill
Venables
"Strategies for Developing and Validating Predictive Models" with Frank
Harrell
"Financial Topics in S-PLUS" with Doug Martin

For current information: www.mathsoft.com/splus/splsprod/userconf.htm
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news