Old News
Lecture Intelligent Data Analysis
Summer Term 2010
News
Last assignment sheet online. Some slides added for association rule visualization.
Summary
- General Information
- Schedule and Rooms
- Lecturers
- Conditions for Certificates (Scheine) and Exams
- Prerequisites
- Slides
- Assignment Sheets
- Additional Material
- Software
- Data Sets
- References
- Links
General Information
This page contains information about the lecture "Intelligent Data Analysis" that is held in summer term 2010 by Dipl.-Inf. Matthias Steinbrecher. This page is updated during the course.
In many areas the overwhelmingly large volumes of data can hardly be assessed manually by a user: Think of market basket data collected in supermarkets. The analysis of this data however is important because they may reveal decisive information about e.g. the customers' buying behavior. Therefore, "intelligent" analysis techniques have to be devised that are capable to extract from illegible date comprehensible information.
In this lecture a multitude of methods are introduced that range from classical statistics (descriptive statistics, inferential statistics, parameter estimation, regression) over association rules, Bayesian classifiers, decision and regression trees, fuzzy data analysis to clustering techniques.
| back to top |
Schedule and Rooms
| Weekday | Time | Room | Begin | |
|---|---|---|---|---|
| Lecture | Wednesday | 09.15-10.45 | 29-K059 | 31.03.2010 |
| Exercise | Tuesday | 11.15-12.45 | 22a-122 | 06.04.2010 |
| back to top |
Lecturers
If you have questions regarding the lecture or exercise, please contact me (via e-mail if possible).
- Matthias Steinbrecher, msteinbr@iws.cs.uni-magdeburg.de
| back to top |
Conditions for Certificates (Scheine) and Exams
Certificate (Übungsschein): There are assignment sheets published every week. Assignments the solutions of which you want to present in the exercise lecture have to be ticked beforehand on a votation sheet that is handed out prior to every exercise lecture. If ticked, you may be asked to present your solution in front of class. The solutions need not necessarily be completely correct, however, it should become obvious that you treated the assignment thoroughly. You are granted the certificate (Schein), if (and only if) you
- ticked at least two thirds of the assignments,
- presented at least two times a solution during the exercise, and
- passed a small colloquium (approx. 10 min) or a written test (if there are more than 20 students) after the course.
Exam: If you intend to finish the course with an exam, your are not required to meet the certificate conditions. However, you are of course encouraged to also solve the assignments. Regarding the exam, please register with the lists lying out at the Prüfungsamt (at the end of the term). The exam has to be announced to the Prüfungsamt two weeks prior to the exam date via this application form. The exam consists of a 20 to 25 minutes oral examination about the subjects presented during the course. Emphasis is put on understanding rather than formal details. The final marks will be in the following range: 1.0, 1.3, 1.7, ...,3.7, 4.0, 5.0. If more than 20 students intend to do an exam, the oral exam may be converted into a written examination.
| back to top |
Prerequisites
You should have background knowledge on fundamentals of computer science such as algorithms, data structures etc. Also, some insights into probability theory are highly recommended.
| back to top |
Slides
The lecture slides can be found below.
| back top top |
Assignment Sheets
The assignment sheets will be published weekly at this location.
| back top top |
Additional Material
Feel free to check out the following supplementary material that augment the lecture and exercise.
- Introduction to Statistics (German)
- Paper on Bayesian classifiers and fuzzy clustering
- Paper on Decision Trees (German)
- Pseudo code of the Apriori Algorithm
- Derivation of polynomial regression (German)
- Chi^2 Test (German)
- Answer Trees (German)
| back to top |
Software
Here you find links to programs with wich simple data analysis tasks can be solved.
- Confidence Interval Visualization
- Multi Polynomal Regression
- Naive and Full Bayesian Classifiers
- Visualization of Bayesian Classifiers
- Induction of Decision and Regression Trees
- Visualization of Decision and Regression Trees
- Training of Neuronal Networks (Multi-Layer Perceptrons)
- Training of Neuronal Networks (Radial Basis Functions)
- Probabilistic and Fuzzy Clustering
- Induction of Association Rules
- C program to compute the quantiles of the normal distribution
| back to top |
Data Sets
Here you find some example data sets that can be fed into the algorithms above.
| back to top |
References
- Intelligent Data Analysis: An Introduction (2. edition)
D. Hand and M. Berthold (eds.)
Springer-Verlag, Berlin, 2002 - Elementare Einführung in die angewandte Statistik
K. Bosch
Vieweg, Wiesbaden, 2000 - Angewandte Statistik (9. Auflage)
L. Sachs
Springer, Berlin, 1999 - Machine Learning
T. Mitchell
McGraw Hill, New York, NY, USA 1997 - Data Mining Techniques. For Marketing, Sales, and Customer Support.
M.J. Berry and G.S. Linoff
J. Wiley & Sons, Chichester, United Kingdom 1997 - Data Mining - Theoretische Aspekte und Anwendungen
G. Nakhaeizadeh (Hrsg.)
Physica-Verlag, Heidelberg, 1998 - Data Mining Methods for Knowledge Discovery
K. Cios, W. Pedrycz, R. Swiniarski
Kluwer, Dordrecht, Netherlands 1998
| back to top |
Links
- Otto-von-Guericke University of Magdeburg
- Faculty of Computer Science
- Department of Knowledge and Language Engineering
- Computational Intelligence Group
- Other Lectures of the Workgroup
| back to top |
