Applied Mathematical Optimization (M. Sc./Ph.D.)
Applied Mathematical Optimization (AMO) - offered by IS3 - covers the use of information systems and quantitative decision-making under uncertainty in a vibrant area: Sustainable smart energy markets. Complementing courses such as Data Science and Machine Learning, and Analytics and Applications, this course focuses on mathematical optimization, especially from the operations management (OM) perspective.
During a brief introduction, you will learn and understand how energy systems change due to renewable energies and what these changes imply for operations management of different stakeholders. You will also learn to apply optimization techniques to answer important questions, for instance: How to schedule the production of power plants in the presence of stochastic renewable energy sources? How to invest in generation facilities to serve future loads with minimum cost and emissions? Given generators and loads, how can we find equilibrium prices?
Energy systems exhibit inherent uncertainty. For instance, solar panels only produce energy in sunny conditions. Managing risks associated with this uncertainty is an increasingly important area of management, not only in the energy domain, but also in finance, health care, marketing, supply chain management, etc. You will learn how to use data-scientific methods to create scenarios that represent potential states of the future. Based on these
scenarios, the stochastic or robust optimization methods you will work with help you find the best line of action - either in the worst case or in expectation.
Beyond stochastic and robust techniques in mathematical programming, we will also discuss bi-level optimization, as well as approaches to decomposition and distributed computing in operations management. These techniques reflect the complex interactions different stakeholders may have in areas where central control is either infeasible or impractical: Your energy supplier can and should not control each of your household devices, rather you might want to control them based on (price) signals. Or when different suppliers, each with their own profit in mind, offer a homogeneous good like electrical energy in an auction - how can prices be formed?
While the presented techniques are advanced, we do not expect pre-existing knowledge in any of the aforementioned areas. An interest in quantitative and mathematical methods is sufficient. After several lecture blocks to jump-start your skill set, we will work hands-on on different problem sets from the energy domain. These use cases could also be an inspiration for potential research during your master thesis.
The grading of the course is based on an individual and a group project (equal weights). There is no exam. During both projects, you are expected to work on hands-on projects that reflect your learnings throughout the semester. Further, you show your ability to persuasively report findings in written and oral form.