General Information

In a master's thesis a student should demonstrate his or her ability to work on a computer science topic understanding, combining, and integrating current lines of research. Topics are usually drawn from research and/or industrial projects. The following pre-requisites emphasize general requirements for work on a thesis. Though, some of them may seem to be obvious, experience is that they are more than often not.

• Good language skills
• Programming skills
• A sound base in mathematics
• Finding and reading relevant literature

My preferences in thesis language and text system used to write the thesis are English and LaTeX, respectively. The final level of scientific quality of the thesis should be such that the main results can be considered for publication. The following list links you to more detailed information on specific topics (feel free to contact me anytime to discuss your interests in a topic and/or own ideas):

Open Topics

• Evolutionary and Reinforcement Learning of Neural Go Players
  
  
• Evolution of Spiking Neural Networks
  
  
• Evolutionary (Non)Linear Equation System Solvers
  
  
• Evolutionary Solutions of Resource Allocation Problems
  
  
• Evolutionary (Non)Linear Feature Construction
  
  

Theses in progress...

• Automatic Sleep Classification in Healthy and Patients Suffering from Disorders of Consciousness Using Permutation Entropy and Artificial Neural Networks
  
  
• Evolution of Artificial Neural Networks Augmented by Neuromodulators for Mobile Autonomous Robots
  
  

Completed Theses

• Approximating an Optimal Strategy for Full-scale Heads-up No-Limit Holdem Poker By Coevolutionary Algorithms, Maximilian Schorr, 2015.
  
  
• LitSOM - Automatic Mapping of Russian Authors Using Self-Organizing Maps, Gernot Howanitz, 2013.
  
  
• Simulation of the Evolution of the Behavior of Migratory and Sedentary Birds, Josef Seibl, 2012.
  
  
• Simulation and Optimization of Vascular Bifurcations by Numerical Solutions of the Navier-Stokes Equations Using Finite Elements, Robert Resch, 2008.
  
  
• Adaptive Autonomous Traffic Control Based on Neural Reinforcement Learning, Andreas Ban, 2008.
  
  
• An Application of Asynchronous Replication in a Loosely Coupled Data Base Cluster, Richard Brunauer, 2008.
  
  
• Computational Intelligence Techniques for Car Crash Detector Optimization, Simon Sigl, 2007.
  
  
• [Prize] Automatic Classification of Geology from Tunnel Boring Machine Data, Gerhard Mitterlechner, 2007.
  
  
• Simulation and Optimization of Vascular Bifurcations by means of Computational Fluid Dynamics, Gerhard Gaube, 2006.
  
  
• Detection of Mental Blocks in Humans Based on Physiological Measures, Franz-Josef Auernigg, 2006.
  
  
• Hebbian Reinforcement Learning in Artificial Neurocontrollers, Thomas Mirlacher, 2005.
  
  
• [Paper] Evolution of Behavior Based Fuzzy Controllers, Markus Spitzlinger, 2005.
  
  
• Analysis of Diabetes Time Series and Blood Glucose Prediction with Artificial Neural Networks, Martin Stockhammer, 2004.
  
  
• [Paper] Evolution and Coevolution of Artificial Neural Networks Playing Go, Peter Maier, 2004.
  
  
• [Paper] A Framework for Solving Resource Allocation Problems Driven by a Modular Evolutionary Engine, Oskar Preinfalk, 2004.
  
  
• [Prize] Automated Quality Control of Powder Metallurgic Parts Based on Statistical Pattern Recognition Using Finite Mixtures, Karl Fürlinger, 2001.
  The thesis has been awarded with prizes by Miba AG and by the Austrian Science Ministry.
  
  
• An Evolutionary System Optimizing Orchestra Staff Schedules, Adrian Poje, 1999.
  
  
• A Comparison of Selection Methods in Genetic Algorithms Based on a Set of Selection Metrics, Thomas Schell, 1998.
  
  

Helmut A. Mayer
helmut@cosy.sbg.ac.at