Tik-122.102 Special Course in Information Technology VI (L V)
|Lecturers:||Prof. Heikki Mannila, D.Sc. Jaakko Hollmén|
|Assistant:||M.Sc. Jouni Seppänen|
|Place:||Lecture hall F in the main building|
|Time:||Wednesday, 14:15 -- 16:00, beginning January 17|
|Language:||English (or Finnish)|
Advances in data acquisition and storage have allowed building huge databases containing, e.g., supermarket transactions, telephone call details, images of astronomical objects, and biological data. The discipline of data mining is concerned with converting such large data sets into useful information. The results of data mining can be models of the whole data set or local patterns evident in some parts of the data.
Data mining differs from traditional statistical analysis in that the data sets are huge, often gigabytes or terabytes of size, and that the data is often acquired for purposes other than data mining. For example, supermarkets store customer transactions in a database primarily for reasons related to bookkeeping, and the data miner should find out whether the data has some interesting properties that would make it useful otherwise.
This seminar course is based on the book
Hand, Mannila, Smyth: Principles of Data Mining,soon to be published by MIT Press. The book covers a broad set of topics such as methods of exploratory data analysis, visualization, dealing with uncertainty, applying optimization methods, clustering, classification, regression, data organization, finding patterns and rules, and retrieval by content.
The course is arranged in a seminar form, and it is aimed at graduate and advanced undergraduate students. To pass, you must give one 45-minute oral presentation, write a summary for the others, and participate actively in the other presentations. If you find some topic especially interesting, it is possible to do also a special assignment or thesis.
To enroll, use the Topi system.
Please feel free to contact the lecturers for further information on the course, or if you would like to do a special assignment or thesis on a data mining topic.
Jaakko Hollmén, B234, Jaakko.Hollmen@hut.fi, tel. 451 5290
Heikki Mannila, B246, Heikki.Mannila@hut.fi, tel. 451 4852
Jouni Seppänen, B246, Jouni.Seppanen@hut.fi, tel. 451 4852