CSE 572 DATA MINING

(August 21 - December 5, Fall 2006)

We will together hear, see and do in this class in many forms including lectures, invited talks, discussions, research paper reading assignment, a project, and presentation, in addition to homework, quizzes, and exam(s).  We will create opportunities to learn from each other. If you find anything interesting about data mining, please share with all.

Your suggestions are most welcome. Please send email to huan.liu at asu.edu

OUTLINE

1. Introduction to Data Mining
2. Classification Methods (ensemble methods, skewed data, cost-sensitive classification)
3. A Brief Review of Probability and Entropy
4. Performance Evaluation (Measures, Comparison between two algorithms)
5. Data, Data Preparation, and Data Preprocessing (feature selection, discretization, sampling, instance selection)
   
6. Clustering Methods (subspace clustering, CLIQUE)
   
7. Association Rules
8. Current Challenges
1. USuggest  - an invited talk on data mining in a Web application (5:15pm Oct 2, 2006)
   
9. Some Thought-Provoking Applications (steganography and steganalysis, streaming data extraction, gene selection)
   

ASSIGNMENTS (To be updated)

1. There is an extra credit (up to 10%) for class participation and discussion.
2. Assignments 1 and 2 are stated in the first set of slides.
   
3. Research paper reading assignment, and selected presentations (We will discuss more in class)
   
4. Everyone will need to work on a course project. Thinking of a suitable project will inspire your learning, doing the project will allow you to apply what you learn, and more importantly, explore beyond what is taught in class.  (Project Categories)
   
1. Project proposal
2. Progress report
3. Project presentation  [please submit the URL links to TA], schedule will be available at myASU
• Every student will have 15 minutes to present his project including 2 minutes for Q/A
• 12-15 slides will be usually sufficient for this type of presentation
  
5. Project final report and demo
   

EXAMS

• Exam 1 on Oct 23 Monday, closed book, in class. A review will be held on Oct 18, Wednesday.
  

LINKS

• KDNuggets - a comprehensive site for many resources of KDD
• KDD Cup 2001 - included here as an example, you can find many new ones at SIGKDD conference sites.
• Tutorial on Feature Selection for Data Mining
  
• Tutorial on Mining and Indexing of Time Series Data under Research Information
• Data mining software packages recommended by Chris Mayer
• Random Forests by Leo Breiman
• Bioinformatics Journal, Oxford University Press, available to ASU students
• FAQ's