Monday, 7 April 2025 @ 14:00–15:00 CEST

On-site:

University of Vienna
Seminarraum 2 (1. UG)
Währinger Straße 29
1090 Vienna

Online:

https://univienna.zoom.us/j/67032386717?pwd=g8HOG2oRrWK6T5cvmRA7bv17QRzq72.1

Meeting ID: 670 3238 6717
Passcode: 440328

Learning meets Operations Research

Abstract
:

The integration of machine learning into methods for solving combinatorial optimisation problems is a rapidly growing area of research. A key question is how to use learning in operations research algorithms, since combinatorial optimisation problems are different from most problems successfully solved by machine learning.

While recent research has mainly focused on the integration into heuristic algorithms, this talk deals with the integration of machine learning into an exact method. More specifically, we discuss the applications in branch-and-price algorithms for routing problems while maintaining the exactness of the algorithm. The central hypothesis is that algorithmic components can be configured by learning from previously solved problem instances to make better decisions.

Bio
:

Christian Tilk is Assistant Professor at the Department of Business Decisions and Analytics at the University of Vienna.  He received his PhD from the Johannes Gutenberg University Mainz in 2016. His research interests cover all areas of Operations Research. Current topics are modelling and solving complex vehicle routing problems, city logistics, predictive maintenance in railway traffic, and the application of machine learning to enhance operations research methods.