Welcome!%20MSCIT%20521:%20Knowledge%20Discovery%20and%20Data%20Mining

1
Welcome! MSCIT 521 Knowledge Discovery and
Data Mining

• Qiang Yang
• Hong Kong University of Science and Technology
• qyang_at_cs.ust.hk
• http//www.cs.ust.hk

2
Data Mining An Example

• KDDCUP from past years
• 2007
• Predict if a user is going to rate a movie?
• Predict how many users are going to rate a movie?
• 2006
• Predict if a patient has cancer from medical
  images
• 2005
• Given a web query (Apple), predict the
  categories (IT, Food)
• 1998
• Given a person, predict if this person is going
  to donate money

• In general, we wish to
• Input Data
• Output
• Build model
• Apply model to future data

2
3
Data Mining Convergence of Three Technologies
3
4
Definition Predictive Model

• A black box that makes predictions about the
  future based on information from the past and
  present
• Large number of inputs usually available

4
5
How are Models Built and Used?

• High Level View

5
6
The Data Mining Process
6
7
What does the Real World Look Like
7
8
Predictive Models are

• Decision Trees
• Nearest Neighbor Classification
• Neural Networks
• Rule Induction
• Clustering

8
9
Course Description

• Data Mining and Knowledge Discovery
• Focus
• Focus 1 Theoretical foundations in Pattern
  Recognition and Machine Learning
• Algorithms
• Differences?
• where they apply?
• Focus 2 Broad survey of recent research
• Focus 3 Hands-on, apply algorithms to KDD data
  sets

10
Topic 1 Foundations

• Classification algorithms
• Clustering algorithms
• Association algorithms
• Sequential Data Mining
• Novel Applications
• Web
• Customer Relationship Management
• Biological Data

11
Topic 2 Hands On

• Apply learned algorithms to selected data sets
• Homework assignments
• Get familiar with existing system packages and
  libraries
• In-class workshops
• Programming Assignments

12
Important Sites

• Instructor Web Site
• http//www.cse.ust.hk/qyang/521
• TA Kaixiang Mo
• Assignment Hand-in online
• csit5210_at_ust.hk
• Course Discussion Site
• Check out the web cite

13
Prerequisites

• Statistics and Probability would help,
• but not necessary
• Pattern Recognition would help,
• but not necessary
• Databases
• Knowledge of SQL and relational algebra
• But not necessary
• One programming language
• One of Java, C, Perl, Matlab, etc.
• Will need to read Java Library

14
Grading

• Grade Distribution
• Assignments 20
• Course Project 20
• Exams 60
• Midterm 20
• Final 40

15
More info

• Textbooks For reference only
• Introduction to Data Mining by Pang-Ning Tan,
  Michael Steinbach, and Vipin Kumar, Pearson
  International Edition, 2005.
• Data Mining.  by Ian Witten and Ebe Frank.
  (Google books)
• Data Mining -- Concepts and Techniques by Jiawei
  Han and Micheline Kamber. Morgan Kaufmann
  Publishers.
• Available in our bookstore