Seminar: Discrete Optimization and Machine Learning


Prof. Dr. Andreas S. Schulz, Diogo Poças, PhD

Brief description

The seminar will focus on recent research at the interplay of machine learning and discrete optimization. On the one hand, discrete optimization problems are becoming increasingly important in machine learning. We will identify some of these problems as well as structures that make it possible to efficiently obtain exact or approximate solutions even at very large scales. On the other hand, well-established solution methods for NP-hard discrete optimization problems may benefit from the application of learning techniques. We will study how machine learning can improve the design of exact and heuristic algorithms in discrete optimization. Students are expected to read and thoroughly understand original research papers, and to deliver an oral presentation.

  • A list of original research papers will be handed out shortly before the start of the term.
  • Students are required to pick a paper during the first week of classes.
  • Students will give oral presentations on their papers in December.
  • All presentations have to be given in English.



  • Discrete Optimization (MA3502) or Combinatorial Optimization (MA4502).
  • Machine Learning (IN2064) or Statistical Modeling and Machine Learning (IN2332).