Teaching - Postgraduate
Course material for ID5059 Knowledge Discovery and Data Mining, Semester 2 of 2012-13. The coursework was co-developed with Carl Donovan.
Lecture notes
- Lecture 01 - Introduction - Slides - Notes - Breiman paper
- Lecture 02 - Basis Functions - Slides - Notes - Kondor paper
- Lecture 04 - Model Fit Measures - Slides - Notes - Stokes paper
- Lecture 05 - Model Selection - Slides - Notes - Efron paper
- Lecture 06 - Tree-based Methods - Slides - Notes - Wang et al. paper
- Lecture 07 - Regression and Decision Trees - Slides - Notes
- Lecture 08 - Regression Trees - Slides - Notes
- Lecture 09 - Classification Trees - Slides - Notes
- Lecture 10 - Complexity & Numerics - Slides - Notes
- Lecture 11 - Neural Nets - Slides - Notes
- Lecture 12 - Neural Nets - Slides - Notes
- Lecture 13 - Bayesian Classification - Slides - Notes
- Lecture 14 - Classification Evaluation & ROC - Slides - Notes
- Lecture 15 - ROC, AUC & Lift- Slides - Notes
- Lecture 16 - Nonparametric Bootstrapping - Slides - Notes
- Lecture 17 - Bagging - Slides - Notes
- Lecture 18 - Boosting - Slides - Notes - Freund & Schapire paper
- Lecture 19 - Clustering and Random Forests - Slides - Notes
Practicals
- Practical 01 - Auto MPG - Spec. - Resources - Data
- Practical 02 - Kickers - Specification, due dates and tips.
Tutorials
- Tutorial 01 - Titanic Survival - zip file
- Tutorial 02 - Regression Trees - zip file
- Tutorial 03 - Model Evaluation using ROC, AUC and Lift - zip file. These examples are taken from Andrew Ziem's blog.
The Elements of Statistical Learning by Hastie, Tibshirani & Friedman is available from Stanford University. There are other useful resources at the same location, including sample R files.
The Titanic tutorial material is available as a zip file.
R and Data Mining: Examples and Case Studies by Yanchang Zhao. This is the PDF of a textbook containing many worked examples (in R) together with detailed explanations of the theory and the the technicalities involved.
R material for non-geeks is available from the University of California at Davis.
As this appears not to be in the old examp paper repository, I've made available the 2010-20 MT5759 exam. Warning! Both the structure and content are likely to change this year; this document gives you some insight into the type of question set for Maths masters students.