TIES443. Introduction to Data Mining.

Introduction

Today, to be successful, business organizations need rapid and easy access to information about their customers, internal finances, and external market conditions—information collectively known as Business Intelligence (BI). Modern business environment requires a timely and accurate insight into up-to-date business conditions. Without relevant information extracted from data decisions may be more risky and may not be the best one for the faced challenges.
This course provides basic introduction into key areas such as OLAP (that stands for On Line Analytical Processing) Design, Data Warehousing (DW), and Data Mining (DM) emphasizing DM to ensure that students of the course will gain the necessary background and skills to turn available data into valuable and useful information.
The course consists of lectures, tutorials (demonstrations) followed by laboratory works, and seminars where students are required to demonstrate their group work.

Objectives of the Course

Provide basic introduction into key areas such as
- OLAP (that stands for On Line Analytical Processing) Design,
- Data Warehousing (DW), and
- Data Mining (DM)
Provide an overview of most common tasks and application areas of DM

Prediction and knowledge discovery

Provide an overview of most common techniques used in DM

Building and evaluating predictive and descriptive models

Ensure that students of the course will gain the necessary background and skills

to turn available data into valuable and useful information

Overview of the Course

Lectures 15*2 = 30 hours
- Wed 8:15 – 10:00, Thu 12:15 – 14:00, Fri 10:15 – 12:00
- all in Ag Beeta
- Nov 17, 12.00 – 14.00 Sami’s public examination of PhD (after the lecture)
Tutorial followed by an assignment 5*2 = 10 hours
- Tue 14.15 – 16.00, Ag B212.2 (Mountains),
- but week 50: Wed 8:15 – 10:00
Seminar 2 hours
- Ag B212.2 (Mountains)
- 5-10 min presentation by each student about the final Assignment
Final assignment (no final exam)
- to be sent to mpechen{AT}cs{dot}jyu{dot}fi and samiayr{AT}mit{dot}jyu{dot}fi by the end of Jan’07 (always use TIES443 keyword in the subject field)

Instructors

Mykola Pechenizkiy	Sami Äyrämö
E-mail: mpechen{AT}cs{dot}jyu{dot}fi office: AgC 414.3 Tel.: 014 260 4907 Home Page	E-mail: sami.ayramo{AT}mit{dot}jyu{dot}fi office: AgC 416.2 Tel.: 014 260 2533 Home Page

Lectures & Handouts

Date and Time	Lecture title	Instructor	Handouts
Wed, 1.11.06, 8:15-10:00	Introduction to the course	Mykola Pechenizkiy	Handouts
Thu, 2.11.06, 12:15-14:00	Business Intelligence	Mykola Pechenizkiy	Handouts
Fri, 3.11.06, 10:15-12:00	Data Warehousing	Mykola Pechenizkiy	Handouts
Wed, 8.11.06, 08:15-10:00	Data Mining: basic concepts	Sami Äyrämö	Handouts
Thu, 9.11.06, 12:15-14:00	Data Mining process	Sami Äyrämö	Handouts
Fri, 10.11.06, 10:15-12:00	Data preprocessing	Sami Äyrämö	Han & Kamber's book: Chapt.3: "Data Preparation" Tan et al. book: Topic 2: Data
Wed, 15.11.06	Clustering. Part I	Sami Äyrämö	Han & Kamber's book: Chapter 8 Tan et al. book: Present. 8
Thu, 16.11.06	Clustering. Part II	Sami Äyrämö	Tan et al. book: Presentation 9
Fri, 17.11.06	Visualization	Mykola Pechenizkiy	Handouts
Wed, 22.11.06	Classification. Part I	Mykola Pechenizkiy	Handouts
Thu, 23.11.06	Classification. Part II	Mykola Pechenizkiy	request by e-mail
Fri, 24.11.06	Evaluation	Mykola Pechenizkiy	Piatetsky-Shapiro & Parker course: Modules DM10-DM11
Wed, 29.11.06	Association rules	Mykola Pechenizkiy	Tan et al. book: Present. 6-7 Piatetsky-Shapiro & Parker course: Beginning of module DM9
Thu, 30.11.06	Advanced and Miscellaneous issues	Mykola Pechenizkiy	Handouts
Fri, 1.12.06	Closing lecture	Sami Äyrämö	Handouts

Tutorials & Assignments

Date and Time	Title of Tutorial	Instructor	Handouts	Assignment
Tue, 7.11.06, 14:15-16:00	Prototyping DM techniques with WEKA and YALE open-source software	Mykola Pechenizkiy	Handouts	Assignment
Tue, 14.11.06, 14:15-16:00	Prototyping DM techniques with Matlab environment	Sami Äyrämö	Handouts	Iris dataset
Tue, 21.11.06, 14:15-16:00	Mining time-series data	Mykola Pechenizkiy	request by e-mail	Assignment
Tue, 28.11.06, 14:15-16:00	Mining image data	Sami Äyrämö	Handouts	Assignment
Wed, 13.12.06, 08:15-10:00	Mining textual data	Miika Nurminen	Slides and reading	Assignment

Seminar & Final Assignment

Start to Prepare for the Final Assignment
Seminar will be on Tuesday, Nov 28, right after tutorial (16.15-18.00)

Decide on application domain that interests you
Two types of assignments: (1) mining real dataset, and (2) comparison of techniques
Search for a dataset (e.g. from KDD Cups) if you decide to analyze data
Search for a collection of datasets if you decide to compare DM techniques
Define the problem: (1) goal, (2) input, (3) constraints or background
5-10 min presentation
Mykola and Sami will give you feedback
You will try to do what you promise during the seminar, report you work

There will be additional summary information placed to the course webpage after the seminar if needed.

Literature

Introduction to the field

Witten I., Frank E. 2000. Data Mining: Practical machine learning tools with Java implementations", Morgan Kaufmann, San Francisco. (book & software page)
Crawford D. 1996 Special Issue on Data Mining. Communications of the ACM, Volume 39, Number 11, November, 1996
Reinartz, T. 1999, Focusing Solutions for Data Mining. LNAI 1623, Berlin Heidelberg.
Han J. and Kamber M. 2000, Data Mining: Concepts and Techniques, The Morgan Kaufmann Series in Data Management Systems, Morgan Kaufmann Publishers, 550 pages. ISBN 1-55860-489-8 (ppt slides to the book)
Data Mining: A Practitioner’s Approach, ELCA Informatique SA, 2001
CRISP-DM 1.0: Step-by-step data mining guide, SPSS Inc.

Mykola Pechenizkiy