ECE 594N: DATA MINING

Instructor: Manjunath, manj@ece.ucsb.edu

Data mining refers to tools and techniques for processing and managing large collections of data with the main objective being able to detect significant "patterns" or associations in such data sets. As such, it has a wide range of applications to problems in natural and social sciences, medicine, finance, and marketing. This introductory course will cover some of the basic principles of data mining with emphasis on data mining tasks and algorithms. These will include, for example, tools for classification and clustering, data structures for organizing high-dimensional data, association rules for mining, and retrieval by content (mostly chapters 9-14 in the recommended text above, but I will also be using other resources.

Recommended Book(s):

Principles of Data Mining by Hand, Mannila and Smith, MIT Press, 2001. A good introductory book on data mining, with an emphasis on the tools and techniques.

Data Mining: Concepts and Techniques by Han and Kamber, Morgan Kauffman, 2001. another well written book, from a database point of view. The author has slides (ppt) available on his web site for each of the chapters. http://www.cs.sfu.ca/~han/dmbook

In preparing the lecture notes, I have made (extensive) use of content from several sources on the web, including the following course web sites:

Padhraic Smyth, UC Irvine, course on Data Mining, http://www.ics.uci.edu/~smyth/courses/ics278/
Osmar Zaiane, University of Alberta, course on Knowledge Discovery in Data, http://www.cs.ualberta.ca/~zaiane/courses/cmput695/
Christos Faloutsos, CMU, course on Multimedia Databases and Data Mining, http://www-2.cs.cmu.edu/~christos/courses/826.S02/
Tom Minka, CMU, course on Data Mining, http://www.stat.cmu.edu/~minka/courses/36-350/ (note that his 2001 course has the links to lecture slides, http://www.stat.cmu.edu/~minka/courses/36-350.2001/)

The above course web pages should also provide you with a wealth of information on course projects etc. that you should "mine" for getting ideas for your own term projects.

Lecture 1 (03/31/2003): Introduction. (acknowledgments: zaiane, minka, and others)

Lecture 2 (04/2/2003): Data Warehouses (chapter 2 slides from Han).

H/W #1: Due April 10. See the Lecture #2 slides for details.
An Overview of Data Warehouse and OLAP Technology, Chaunduri, Surajit and Dayal, Umeshwar; March 1997; ACM Sigmod Record; Provides a good introduction to data warehouses and OLAP that we discussed in class. (local copy here).

Lecture 3/4 (04/7,9/2003): Data Preprocessing (chapter 3 from Han)

Classification Methods

Lecture 5 (04/14/2003) [Final Version 4/14] An Overview of Data Mining Algorithms (Chapters 5, 6, 7 from Hand's book).

Lecture 6 (04/16/2003) [Final Version 4/16] Classification Part I (LDA and CART) (Ch 10 (Hand), Ch 7 (Han))

Lectures 7, 8 (04/21,22/2003) [Final Vesion 4/23] Classification Part II (Ch 10 (Hand), Ch 7 (Han)) [a handout on backpropagation learning will be distributed in class on Monday, 4/28]

Applications: Additional reading
- Sahami, M., Dumais, S., Heckerman, D., and Horvitz, E. 1998. A Bayesian Approach to Filtering Junk E-Mail. In Learning for Text Categorization: Papers from the 1998 Workshop. AAAI Technical Report WS-98-05.
- Jason T. L. Wang, Qicheng Ma, Dennis Shasha and Cathy H. Wu. "New Techniques for Extracting Features from Protein Sequences", IBM Systems Journal, Special Issue on Deep Computing for Life Sciences, Vol. 40, No. 2, 2001.

Clustering Methods

Lecture 9, 10, 14 (04/28, 30/2003) [final 5/14] Clustering Methods part I

Lecture 14 (04/14/2003) [final 5/14] Clustering Methods part II

Reading

Tian Zhang, Raghu Ramakrishnan, and Miron Livny. BIRCH: An Efficient Data Clustering Method for Very Large Databases. In Proceedings of the 1996 ACM SIGMOD International Conference on Management of Data, pages 103--114, Montreal, Canada, 1996. http://citeseer.nj.nec.com/zhang99birch.html
S. Guha, R Rastogi, and K. Shim. CURE: An efficient clustering algorithm for large databases. In Proceedings of ACM SIGMOD International Conference on Management of Data, pages 73--84, New York, 1998. ACM. http://citeseer.nj.nec.com/guha98cure.html
Sudipto Guha, Rajeev Rastogi, and Kyuseok Shim. ROCK: a robust clustering algorithm for categorical attributes. In Proc. of the 15th Int'l Conf. on Data Eng., 1999. http://citeseer.nj.nec.com/guha00rock.html
Martin Ester, Hans-Peter Kriegel, Jörg Sander, Xiaowei Xu, (DBSCAN) A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise, Proceedings of 2nd International Conference on Knowledge Discovery and Data Mining (KDD-96). [PDF]
M. Ankerst, M. Breunig, H.-P. Kriegel, and J. Sander. Optics: Ordering points to identify the clustering structure, SIGMOD'99. [PDF]
G. Sheikholeslami, S. Chatterjee, and A. Zhang. WaveCluster: A multi-resolution clustering approach for very large spatial databases. VLDB’98. [PDF]
A tutorial on EM algorithm by Bilmes.

Paper Presentations

Lecture 11 (05/05/2003): Siddiqi, El Saban, Aghili, Sahin
Lecture 12 (05/07/2003): Qamra, Bulut
Lecture 13 (05/12/2003): Byun, Panda, Sritsan Pallavaram, Chen
Lecture 14 (05/14/2003): Niu

Association Mining

Lecture 15: edited slides from Chapter 6, Han's book. [05/19/2003, draft PDF]

Multimedia Mining

Lecture 16 (05/22/2003): Content based access to images/video [draft PDF]
Lecture 17 (05/26/2003): holiday (memorial day/05/26)
Lecture 18 (05/28/2003): Relevance Feedback [draft PDF, same as L16]

Yong Rui, Thomas S. Huang, and Sharad Mehrotra. Content-based image retrieval with relevance feedback in MARS. In Proc. IEEE Int. Conf. on Image Proc., 1997. http://citeseer.nj.nec.com/rui97contentbased.html

Flashed With FrontFX FrontFlash For Frontpage By XZAKT Media.