Present and past thesis projects under my supervision:
Nikolaus Mayer: Globale Explorationsstrategien (German) This thesis analyzes the effects of globally directed exploration on a learning agent's performance. We introduce the BELT growing-batch algorithm which, for the first time, employs a learning procedure to create an independent policy for exploration itself. Compared to undirected and locally directed exploration, we find global exploration to accelerate the learning of good policies. Bachelor Thesis, finished 09/2011, PDF |
Student | Title | Completion | Type |
---|---|---|---|
Arne Voigtländer | Deep Reinforcement Learning on Real Systems | finished 07/2010 | Master Thesis |
Andreas Witsch (Uni Kassel) | Policy Gradient Methods | finished 06/2010 | Diplomarbeit |
Tobias Springenberg | Machine Learning on Massively Parallel Architectures - A Case Study | finished 09/2009 | Bachelor Thesis |
Philipp Blohm | Beurteilung von Stärken und Schwächen von neuroevolutionären Verfahren zum Einsatz im Bereich des Reinforcement Learnings | finished 10/2008 | Bachelor Thesis |
Christian Müller | Reinforcement-Lernen mit Pac-Man | finished 03/2007 | Bachelor Thesis |
Sebastian Schnock | Implementation und Evaluation von Neuroevolution in CLS | finished 07/2007 | Bachelor Thesis |
Hussain Iqbal | Evolutionary Algorithm using neuro-evolutionary techniques in dynamic environment | finished 08/2007 | Master Thesis |
Michael Wiegand (FH Karlsruhe) | Vorhersage der Bewegung eines autonomen Roboters | finished 03/2007 | externe Diplomarbeit |
Tim Kitzmann | View-Based Object Recognition: A Biologically Motivated Approach using iGRLVQ | finished 12/2006 | Bachelor Thesis |
Andrea Albrecht | Erkennung und Klassifizierung von Objekten mit Neuronalen Netzen | finished 09/2007 | Bachelor Thesis |
Srinivas Gudipally | Survey: Approaches of Artificial Intelligence with the focus on Machine Learning within the RoboCup domain | finished 8/2006 | Master Thesis |
Andreas G. Nie | RoboPECS - Autonome Fußballfans im RoboCup Entwicklung eines Multiagentensystems zur Simulation von Fußballfans | finished 7/2005 | Magisterarbeit |