CSEngMtCpEng 404 Data Mining

1
CS/EngMt/CpEng 404Data Mining Knowledge
Discovery

• Daniel C. St. Clair, PhD Christopher Merz, PhD
• University of MO Rolla Mastercard International
• Lect 1b Intro. to CRISP-DM

2
Joe (Data) Miner
3
Lecture 1 Contents

• Intro to CS/EMgt/CpE 404
• What is data mining KD?
• Data sources
• Data mining tasks
• Introduction to CRISP-DM

DSC
CM
4
What is CRISP-DM?

• The Cross Industry Standard Process for Data
  Mining
• Consortium Members include
• SPSS
• Teradata
• DaimlerChrysler
• URL www.crisp-dm.org
• Product CRISP-DM version 1.0

5
Advantages of CRISP-DM?

• Industry Neutral
• Tool Neutral
• Closely Related to KDD Process Model
• Anchors the Data Mining Process
• Attach data mining goals to business or
  scientific goals
• Prevents requirements drift
• Follow through to deployment

6
Disadvantages of CRISP-DM?

• Less emphasis on addressing scientific problems
• Assumes knowledge of tools and / or modeling
  methods
• Does not fit all problems well

7
How Will We Use CRISP-DM?

• See modified template on class web site
  CRISP-DM-UMR-template.doc
• Required for class project
• Major sections correspond to class project
  milestones
• Class schedule specifies when each major section
  will be discussed

8
The CRISP-DM Process Model
9
The Knowledge Discovery Process
Source Fayyad, U., Piatetsky-Shapiro, G.,
Smyth, P, From Data Mining To Knowledge Discovery
In Databases, AI Magazine, Fall 1996.
10
Relating the CRISP-DM Process to the Knowledge
Discovery Process
6. Evaluation
7. Deployment
5. Modeling
4.3 Construct Data
4.2 Clean Data
4.1 Select Data
3. Data Understanding
2. Business Understanding
11
Relating the CRISP-DM and KDD Process Models

• CRISP-DM subsumes KDD Process
• Up front anchoring of data mining goals to
  business / scientific goals
• Tail end emphasis on deployment
• More emphasis on Data Understanding

12
Peruse CRISP-DM Template

• Section 1 contains guidelines for completing
  document

13
Lecture 2 Covers

• CRISP-DM Section 2 - Business / Scientific
  Understanding
• Data Mining project description

14
Assignment for Week 2

• CRISP-DM Class Template
• Retrieve template
• Display hidden text
• Review Sections 1 and 2
• CRISP-DM Reference
• Visit www.crisp-dm.org
• Retrieve CRISP-DM 1.0 Reference Guide
• NOTE Section numbers do not line up

15
CS 404 Class Information
Instructors Daniel C. St. Clair, PhD Christopher
Merz, PhD University of MO Rolla Mastercard
International Phone [phone] Phone
[phone] e-mail stclair_at_umr.edu e-mail
merzc_at_umr.edu CS 404 web page www.umr.edu/s
tclair or http//web.umr.edu/stclair/class/cl
assfiles/cs404_ws04/
16
CS/EngMt/CpEng 404Data Mining Knowledge
Discovery

• Daniel C. St. Clair, PhD Christopher Merz, PhD
• University of MO Rolla Mastercard International
• Lect 1 Intro. to Data Mining