Here are some informations about the projects (past and present) in my research group:
|
Computer vision for medical endoscopy. New methods for image mosaicing and 3D navigation. |
|||
.png "Connect-4 board. Red to move. Red is under zugzwang and will loose the game. The numbers indicate the states of some so-called \"n-tuples\".")
|
Can a computer learn the complex strategies of board games just from self-play, without explicit programming or exhaustive search? We study games like Connect-4 or Dots-and-Boxes with reinforcement learning methods (TD-learning, TCL), online adaptive learning and n-tuple feature coding. |
|||
|
Recognition of gestures with the help of Wiimote or Kinect. Learning methods: SFA, Hidden Markov, Random Forest. |
|||
|
Constrained optimization: Model-based optimization for nonlinear and restricted prodution processes. BMWi-funded project 2014-2016. |
|||
|
Systematic optimization, especially for machine learning and data mining. BMBF-funded project 2009-2013. |
|||
|
|
