Lehr- und Forschungseinheit für Datenbanksysteme

Breadcrumb Navigation


Knowledge Discovery in Databases I (WS 2021/22)


[28.09.21] All information regarding the lecture are exclusively provided via Uni2Work!!

[28.09.21] Registration via Uni2Work is now open


Umfang: 3+2 Semesterwochenstunden
Dozent: Prof. Dr. Thomas Seidl
Vorkenntnisse: Algorithmen und Datenstrukturen empfohlen, Datenbanksysteme I vorteilhaft.

Anmeldung: über Uni2Work
Übungsleitung: Sandra Obermeier, Janina Sontheim

LMUCast: vorraussichtlich: LMUCast Playlist

Termine und Ort

Vorlesung Di, 9:15 - 11:45 Uhr siehe Uni2Work 19.10.2021
Übung 1 Do, 12:00 - 14:00 Uhr Theresienstr. 39, B 134 28.10.2021
Übung 2 Do, 14:00 - 16:00 Uhr Theresienstr. 39, B 134 28.10.2021
Übung 3 Fr, 12:00 - 14:00 Uhr Online  29.10.2021
Übung 4 Fr, 14:00 - 16:00 Uhr Online 29.10.2021


The vast increase in data volume on almost every field results in increased difficulty or even impossibility for information analysis. Especially in areas as biological measurement evaluation (e.g. gene sequencing, micro-array processes …) or data transaction in large telecommunications or network operators, using data without computational aid is inconceivable. The research area “Knowledge Discovery in Databases (KDD)” investigates solutions to these problems. It combines statistics, machine learning, database systems, and (semi-) automatic extraction methods for valid, new, and potentially useful knowledge from large databases. The term data mining in this context refers to the fundamental step in the KDD process, in which the actual analysis of the data is carried out. Data mining is often applied to large amounts of operational data that are managed separately in so-called data warehouses. The frequently used term Business Intelligence describes, among other things, the application of data mining algorithms to the information provided by a data warehouse in order to support targeted decision-making processes. The lecture gives an overview of the basics of the most important KDD techniques. Particularly: Classification, regression/trend detection, clustering, outlier detection, association rules, and process mining.

To deepen the lecture, exercises are offered in which the presented procedures are further explained and illustrated with practical examples.