Research

Systems optimization is—tightly linked to modelling and simulation—a central issue in solving scientific and engineering problems using computers. In this domain, several core challenges emerge, including multiple competing criteria, large search spaces, mixed discrete-continuous models as well as uncertainty and robustness, and randomized search algorithms like evolutionary algorithms and simulated annealing have become indispensable instruments for tackling them. Since these techniques allow to treat the objective functions as a black box, they are ideally suited to model-oriented problem solving where the mathematical modelling process is the center of attention, in contrast to an algorithm-oriented approach which mainly aims at designing an exact algorithm, possibly at the cost of model simplifications.

Our overall goal is to develop novel concepts, methods, and tools for large-scale optimization in order to improve our ability to analyze and design complex systems. Our research focuses on multiobjective optimization and randomized search algorithms from three different angles: (i) theoretical foundations, (ii) algorithm design and implementation, and (iii) applied decision support in systems biology and engineering design.

Topics

• Black box optimization and randomized search algorithms
• Multiple criteria optimization and decision making
• Biologically inspired computation
• Computational biology and bioinformatics
• Engineering Design

More

• Excerpt from the annual report of the Department of Information Technology and Electrical Engineering
• Introductory material can found in the "Downloads/Materials" section