Tutorial Background Reading:
ICS 278: Data Mining
Spring 2006
Below I will provide links to some papers that complement
the
material presented in class and in the text. The links should be
particularly useful for students doing projects on any of the topics
mentioned below.
- ·
Predictive Modeling:
- Strategies and methods for
prediction Greg Ridgeway, 2003. A very nice overview paper
on predictive modeling techniques for both regression and
classification, with an integrated discussion of many well-known
techniques such as logistic regression, nearest-neighbor classifiers,
boosting, trees, SVMs, neural networks, and more.
- Recent
advances in predictive machine learning, Jerome Friedman, Phystat
2003. An informative review of two of the most popular and accurate
techniques in classification and regression, namely, SVMs and boosted
decision trees.
- A
short introduction to boosting Robert Schapire and Yoav Freund,
1999.
- Suport vector
machines: tutorial slides from Andrew Moore (nice visual examples
of how SVMs work)
- Support
vector machines: hype or hallelujah?, summary review article by
Kristin Bennett and Colin Campbell, SIGKDD Explorations, 2000.
- ·
Probabilistic and Bayesian Methods
- Bayesian analysis for data mining,
David Madigan and Greg Ridgeway, 2003. A good overview discussion of
Bayesian methodology with links to data mining.
- Graphical models, Michael Jordan,
Statistical Science, 2004. Broad introduction to the general framework
of graphical models.