Inaugural symposium of the Hertz Chair for Algorithms and Optimization - László Végh breaks new ground between mathematics, computer science and economics
The new Hertz Chair of Algorithms and Optimisation, Lazlo Végh, during his inaugural speech. Finding the best possible solution to complex problems with the help of mathematics, computer science and economics - that is the goal of Professor László Végh, who has been the University of Bonn’s new Hertz Chair for Algorithms and Optimization since August 2024. Together with his transdisciplinary team, he develops efficient algorithms and expands our understanding of the limits of computability. The inaugural symposium on 25 April 2025 provided a broad insight into the various research areas - from game theory to the latest developments in the theory of optimization.
Modern digital applications often require increasingly sophisticated calculation methods in order to deliver optimal results. Linear optimization, also known as linear programming, is a key underlying mathematical model developed back in the 1940s, with a wide range of applications in practice - for example in transport, order, event or personnel deployment planning. While efficient solution methods are already available, the ultimate theoretical complexity of the problem solution, the question of how fast an optimal solution can be found, remains open.
Full house for the inaugural lecture - the audience followed the presentations with rapt attention.
In his inaugural lecture entitled ’The Discrete and Continuous Sides of Linear Optimization’, Professor László Végh gave an overview of classic results and current developments in the field of linear optimization and also presented the long-term goal: the development of so-called strongly polynomial algorithms. With these, the computing time can be limited solely by the number of variables and constraints - regardless of how large the numbers in the model are. A possible step towards calculating the optimal solution. Progress in this area is made possible by combining two central areas of optimization: discrete optimization, which aims to find the best solution from a finite or countable number of possibilities, and continuous optimization, which works with continuous values: Here, the selection for the problem solution must be made from an infinite number of possibilities.
Vera Traub from the ETH Zurich presented her latest ground-breaking result analysing the quality of a famous linear programming relaxation of the Steiner tree problem. This problem is about finding the cheapest network connecting a set of given terminals - especially in chip design.
The subject has fascinated Professor Végh for a long time, he says: "Since my childhood, I have been fascinated by mathematics and by thinking about problems. In linear programming, we are solving geometric problems in high dimensions. I remember first hearing about these problems in high school and trying to wrap my head around it. Ever since then, I have been amazed by the beauty of some of these optimization questions and the combinatorial structures in graphs and networks. These problems are an infinite source of deep and interesting questions. "
Dr Neil Olver from the London School of Economics and Political Science presented improved algorithms for a graph partitioning problem in which a network has to be decomposed into subsets of a given size.
The extended program of the symposium consisted of lectures by invited guests, both international researchers and scientists from the University of Bonn. They provided an overview of the diversity of current research topics in the field and especially at the interfaces between the areas of mathematics, computer science, game theory and economics.
Daniel Dadush from the University of Utrecht during his lecture ’Strongly Polynomial Frame Scaling to High Precision’.
This overarching approach is also important for the university’s Transdisciplinary Research Area (TRA) Modelling, in which Professor Végh’s Hertz Chair is located: ’We hope that the newly appointed Hertz professorship will enable the university and TRA Modelling to forge a stronger link between optimization in mathematics, computer science and economics - in particular through close links with the research strengths already established at the University of Bonn. The aim is to further expand this interdisciplinary area and to work specifically towards future joint initiatives, projects and networking events,’ explain the TRA spokespersons Professor Alexander Effland and Professor Jürgen Gall.
Exchange over coffee and biscuits - lively conversations between colleagues and guests during the break.
The symposium was a successful start to László Végh’s work - full of enthusiasm for the complex questions and problems and anticipation of everything that can come out of it. An inexhaustible source of profound questions and solutions.
