EE 380L : Data Mining (ESE)

Spring 2001
Prof. Joydeep Ghosh
NEWS:
I. Course Information II. References III. Tools
IV. Free Software V. Some Topics VI. Datasets

I. Course information
Information
Course descriptor
Topics/Schedule
TA: Shi Zhong
Email: szhong@ece.utexas.edu
Office Hours: W 8pm-9pm, ENS 518

Project Proposals

Homework

II. References
  1. Reading List 1
  2. Reading List 2 ( Presentation Schedule)
  3. SAS and Enterprise Miner Issues (Where/How to get it, etc.)
  4. Introduction to OLE DB for Data Mining
  5. Citeseer at NEC

III. Tools
  1. Enterprise Miner
  2. UT WNT page w/screenshots
  3. Notes on connecting to published Applications.
  4. Aladdin Expander (.gz tool)
  5. Ghostview (.ps tool)
  6. Acrobat Reader (.pdf tool)

IV. Free Software/Libraries
  • KDNuggets Software

    •  
    Software Libraries (Open-source)
  • BayesBuilder (Win95/NT)
  • DBMiner (Win 95/NT)
  • EXPO SE (Win 95/NT)
    		•
  • MLC++ (C++)
  • TOOLDIAG (C)
  • WEKA (Java)
    		•

    V. Some topics
    Visualization
  • Introduction to Visualization: Vis '96 Tutorial
  • GTM (Generative Topographical Mapping)
  • Hierachical Probabilistic PCA
  • CVIZ Manual - IBM SurfAid
    		•
    Data Reduction
  • Principal Component Analysis (PCA)
  • Synopsis 1
  • Synopsis 2 
  • Notes on Spectral Decomposition
  • MATLAB procedure comparing SVD and PCA
  • Principal Curves (nonlinear)
  • Web-based statistical analyzer
    		•
    Classification
  • ROC curve notes (with figures)
    roc.ps roc.ps.gz roc.zip
  • Support Vector Machines
  • Bayesian Networks
  • Tutorial on Learning Bayesian Networks
  • Bayesian Belief Networks
    		•
    Clustering
  • METIS
    • Miscellaneous
  • EM Algorithm
    		•

    VI. Datasets
    Links (original) Local copy
    1. UCI KDD Archive
    2. UCI Machine Learning databse
    3. Delve
    4. ELENA
    5. PRNN
    6. PROBEN1
    7. StatLib
    8. Statlog
    		 LANS benchmarks Datasets
    KDD Sisyphus I Yes
    KDNuggets Datasets
    Financial Time Series Data
    The Data Mine

    Webmaster: Shi Zhong
    Last updated 01/2001