-----

STATISTICAL LEARNING AND DATA MINING II:
Tools for Tall and Wide Data

State-of-the-Art Statistical Methods for Data Analysis

Harvard Conference Center, Boston, MA: Oct 6-7 (Mon-Tues), 2008.

We are out of office till July 26. You can still send in your registration form, or email Rob Tibshirani with inquiries.


-----

Prediction Surface

A short course given by
Trevor Hastie and Robert Tibshirani
both of Stanford University

This course is the third in a series, and follows our popular past offerings:

Modern Regression and Classification (1996-2000)

Statistical Learning and Data Mining (2001-2005)

This two-day course gives a detailed overview of statistical models for data mining, inference and prediction. With the rapid developments in internet technology, genomics, financial risk modeling, and other high-tech industries, we rely increasingly more on data analysis and statistical models to exploit the vast amounts of data at our fingertips.

In this course we emphasize the tools useful for tackling modern-day data analysis problems. These include gradient boosting, SVMs and kernel methods, random forests, lasso and LARS, ridge regression and GAMs, supervised principal components, and cross-validation. We also present some interesting case studies in a variety of application areas.

This course focuses on both tall data ( N>p, where N is the number of cases, and p the number of features) and wide data (p>N). Typical examples of tall data are credit risk and churn prediction, and email spam filtering. Topics include linear and ridge regression, lasso, and LARS, support vector machines, random forests and boosting. We give in-depth discussion of validation, cross-validation and test set issues.

For wide data, typical examples are gene expression and protein mass spectrometry data, and data from signals and images. Topics include clustering and data visualization, false discovery rates and SAM, regularized logistic regression and discriminant analysis, supervised and unsupervised principal components, support vector machines and the kernel trick, and the careful use of model selection strategies.

The two earlier courses are not a prerequisite for this new course.

The material is based on recent papers by the authors and other researchers, as well as the best selling book:

Elements of Statistical Learning: data mining, inference and prediction

(with J. Friedman, Springer-Verlag, 2001). A copy of this book will be given to all attendees.

The lectures will consist of video-projected presentations and discussion.

  • Further details of the SLDM II course offered in Boston.
  • Registration form for Boston SLDM II course

    http://www-stat.stanford.edu/~hastie/sldm.html