CSI5388: Topics in Machine Learning

Instructor

Nathalie Japkowicz

Office: STE 5029
E-mail: nat@site.uottawa.ca
Telephone: 562-5800 ext. 6693 (Note: e-mail is more reliable)

Meeting Times and Locations

Time:
Wednesday, 1:30pm-4:30pm;
Location: UCU 125

Office Hours and Locations

Times: Mondays and Thursdays, 2:30-3:30
Location: STE 5029

Calendar Description

Machine Learning or data Mining are the areas of Artificial Intelligence concerned with the problem of building computer programs that automatically improve with experience. This course will focus on advanced issues from these fields. Issues such as Feature Selection, Class Imbalances, Cost-sensitivity, One-class learning, Classifier Combination, Performance Evaluation and Visualization (and other topics) will be discussed in depth. Students will be expected to read and critique articles from the recent literature, make two presentations, and propose and complete a research project.

Pre-requesites

CSI5387, although the course can be taken with permission from the instructor (Permission will be granted if the book entitled "Machine Learning" by Tom Mitchell (see below) has been read and well understood prior to taking the course).

Overview

Machine Learning is the area of Artificial Intelligence concerned with the problem of building computer programs that automatically improve with experience. This course will cover, in depth, some advanced topics in the field.

The course will consist of a mixture of regular lectures and student presentations. The regular lectures will cover broad introductions to some of the major areas of research currently under investigation. The student presentations will be of two kinds.

Research Paper Presentations: based on recent research papers that describe new results in the areas discussed in class. The presentations will involve two papers that will need to be contrasted and put in the context of the class discussion.
Theme Presentations: the students will be required to choose a specialised research theme and give an introduction to this theme together with glimpses into recent papers on it. In a way, this will be like teaching a class on the particular theme chosen. The themes that will be covered this term are:
- Genetic Programming
- Evaluating Unsupervised Learning
- Transduction
- Feature Selection for SVM
- Survey of Single Class Learning Methods, Advantages and Disadvantages
- Class Imbalances versus Cost-Sensitive Learning
- Recent Advances in Classifier Combination

Students will be evaluated as follows:

They will have to write short critical commentaries of the assigned research papers on four different weeks (the student decides on the weeks for which s/he will write the commentaries on their own. However, they are not allowed to pick the week of their Research Paper Presentation) [20%],
They will have to give one Research Paper Presentation, as described above [15%]
They will have to give one theme presentation (as described above) [15%]
They will have to propose and carry out the research for their final project. Suggestions for potential projects are given below, but the student is encouraged to pick his/her own topic. Project proposals will be due in mid-semester. [50%]

Topics Covered

Week 1: Review of Machine Learning's main concepts
Slides (1)
Slides (2)
Slides (3)
Week 2: Topics in the Evaluation of Machine Learning algorithms (ROC Curves, Cost Curves) based on: ROC Graphs: Notes and Practical Considerations for Data Mining Researchers, by Tom Fawcett, Submitted to Knowledge Discovery and Data Mining, 2003. [No slides available, at least for the time being' Please, read the Fawcett paper]
Readings:
- What ROC Curves Can't Do (and Cost Curves Can)
- An Empirical Evaluation of Supervised Learning for ROC Area
Week 3: Intro to Genetic Algorithms (Fitness Functions, Genetic Operators, Applications)
Slides

Readings:
- High classification accuracy does not imply effective genetic search
- A Survey of Evolutionary Algorithms for Data Mining and Knowledge Discovery (2001) , Alex A. Freitas.
Week 4: Intro to Unsupervised Learning
Slides

Readings:
- Evaluation Methods for Web Application Clustering
- Iterative Clustering of High Dimensional Text Data Augmented by Local Search
Week 5: Combining Supervised and Unsupervised Learning, Intro to Radial-Basis Functions.
[No available Notes (presentation will be made on the blackboard)]
Readings:
- Learning with labeled and unlabeled data, Matthias Seeger, 2001.
- Understanding the Behavior of Co-training, Kamal Nigam, Rayid Ghani, 2000.
Week 6: Advice on how to formulate a project proposal, presentation by a librarian on ressources available for ML researcher.
Week 7: Topics in Feature Selection (Wrapper, Filter and Hybrid Methods)
Readings:
- Feature Selection for Unsupervised Learning
- Feature Selection for High-Dimensional Data: A Fast Correlation-Based Filter Solution, Yu & Liu, 2003.
Week 8: BREAK
Week 9: Single Class Learning Methods (Autoassociators, single-class SVMs)
Reading:
- Supervised versus Unsupervised Binary-Learning by Feedforward Neural Networks
- Extreme Rebalancing for SVMs: a case study
Week 10: The Class Imbalance Problem
Slides

Readings:
- Mining with Rarity: A Unifying Framework
- A Study of the Behavior of Several Methods for Balancing Machine Learning Training Data
Week 11: Support Vector Machines and Set Covering Machines
Readings:
- Distributional Word Clusters vs. Words for Text Categorization (2002) Ron Bekkerman Ran El-Yaniv Yoad Winter Naftali Tishby.
- Support Vector Machine Learning for Interdependent and Structured Output Spaces
Week 12: Combination of Classifiers (Bagging, Random Forests, Boosting, Error-Correcting Codes, Mixtures-of-Experts) Reading:
- Random Forests [Note: You don't need to buy the paper: U of O has access to the Kluwer Journals on-line. Please find it from the library site.]
- Limited Hierarchical Fusion of Multiple Classifiers for Hyperspectral Data Analysis , Shailesh Kumar and Joydeep Ghosh and Melba M. Crawford, 2002.
Week 13: Project Presentations

Readings

Required

A compilation of selected research papers from the recent literature.

References

Machine Learning, Tom Mitchell, McGraw Hill, 1997.
Introduction to Machine Learning, Nils J. Nilsson (Draft of a Proposed Textbook available on the Web)
Data Mining, Witten, I & Frank, E., Morgan-Kaufmann, 2000.