Thomas Gaertner
Fraunhofer Institut Autonome Intelligente Systeme, Germany

Slides (PDF)


At a time where the amount of data collected day by day far exceeds the human capabilities to extract the knowledge hidden in it, it becomes more and more important to automate the process of learning. The area of computer science concerned with this challenging task is called machine learning and also embraces a wide range of other disciplines including mathematics, engineering, and statistics. This lecture will introduce general machine learning concepts with a particular focus on kernel methods which are currently the most popular and successful class of machine learning algorithms.

While most data collections are huge and contain highly structured information, machine learning research has long assumed that the pieces of information are independent and can be stored as a single entity in a single table of fixed size. The more advanced part of this lecture will concentrate on kernel methods for graphs as these are one of the most popular representations for structured data in mathematics, computer science, engineering disciplines, and other natural sciences. We will focus on algorithmic aspects but also present applications and empirical results related to computational drug design.


Last update December 14, 2006 8:00 EET by local organizers,

© 2006 Institute of Cybernetics at TUT, All Rights Reserved