
5 minute read
About SOR: Stochastic Operations Research
Text: Richard Boucherie
The staff members of the Stochastic Operations Research group (SOR) conduct education and research in applied probability and mathematics of operations research, to contribute to the development of mathematics in a multidisciplinary engineering environment and contribute to a better understanding and functioning of our increasingly complex society. SOR represents the stochastic optimization part of Mathematics of Operations Research within our department. The other parts of MOR are the Discrete Mathematics and Mathematical Programming group headed by Marc Uetz and the Statistics group headed by Johannes Schmidt-Hieber. SOR is partly responsible for the Operations Research and Data Science tracks in the master programme Applied Mathematics. SOR is responsible for module 8 and SOR members teach several courses in other modules of the bachelor programme.
Advertisement
Stochastic operations research aims at the development of mathematical models and methods for the design, control and optimisation of complex systems that are subject to randomness. SOR covers a wide range of topics that is best illustrated by letting all group members pitching themselves.
• Richard Boucherie: I am head of SOR, and responsible for your master programme if you select SOR. My mathematical research interest is queueing theory, in particular queueing networks. Note surprising, I teach the courses queuing theory in module 8 as well as queueing theory in the master programme. I founded the Center for Healthcare Operations Improvement and Research (CHOIR), where I develop mathematical decision solutions that match demand for healthcare with the scarce capacity to provide the right care to the right person at the right place. I also implement these solutions in hospitals to make healthcare healthy.
• Aleida Braaksma: I develop mathematics to optimize healthcare logistics. My dream is that one day patients will be enabled to schedule their own appointments, while the underlying mathematical models guarantee near-optimal schedules for all patients as well as all healthcare providers. From a mathematical perspective, I am thus interested in sequential decision making under uncertainty. I teach Stochastic Dynamic Programing in Module 8 and Markov Decision Theory in the master.
• Jasper Goseling: My research currently focusses on privacy and bias issues arising in artificial intelligence and data science. Using tools from information theory and operations research I develop mathematical frameworks that can be used to obtain insight into problems with current applications, but also to offer key techniques that can be used to resolve these problems. I teach an MSc course on Information Theory and Statistics. Also, I manage the DS&AI Lab of the faculty, through which we offer opportunities for short applied projects with companies in the field of data science and artificial intelligence.
• Marie-Colette van Lieshout: My research interests focus on spatial statistics. Can we quantify the risk of an earthquake in a given region during the coming year? Can we map spatial information in such a way that privacy is respected? Can we relate socio-demographic information to the occurrence of arson fires or burglaries to guide the police's capacity planning? I teach a Master course and supervise student projects on this topic.
• Nelly Litvak: I study complex networks, such as social networks and the World Wide Web. I am interested in design and mathematical analysis of algorithms that help us understand network data, for example, ranking nodes by their importance, finding clusters and communities, and predicting how opinions or epidemics spread on a network. I teach Linear Structures 1 in our Bsc program, and Complex Networks in our MSc program. I love teaching and I always work on new teaching methods that, I hope, help students think and learn deeper. I also wrote two books about mathematics for a broad in Russian (my native language), and co-founded a Facebook group with 30K+ participants about mathematics for non-mathematicians.
• Janusz Meylahn: I'm interested in the dynamics that arise when multiple algorithms learn in the same environment. The difficulty with this topic is that, from the perspective of a single algorithm, the environment changes as the other algorithms learn and vice versa. I teach the honours course on complex networks, a part of the Web Science module and a course on the theory of multi-agent systems in Highlights of OR.
• Jan-Kees van Ommeren: I am interested in developing stochastic models, in particular queueing related models, to real world situations like port logistics or inventory management. Most of the times, the resulting models are too complicated to solve and approximations have to be found. In the Bachelor, I teach and supervise the project Stochastic Simulation; in the Master I teach the course Applied Queueing Models.
• Werner Scheinhardt: Just like you I studied in Twente, both undergraduate and PhD. Most of you know me from the B1 course Probability theory, some also from Markov chains (B2), and a few also from Measure & probability (M1). I enjoy teaching as well as doing research, especially in my favorite field Queueing Theory (which I also teach, but for Mastermath). Queueing is great! Well, at least the theory is.
• Judith Timmer: My research focused on game theory, particularly cooperation under uncertainty. Currently, I coordinate our bachelor programme and teach several courses. For example, Alexander Skopalik and I teach the master course Game Theory. Besides, I am involved in various modules in our bachelor programme.
• Anne Zander: I work on methods to solve sequential decision-making problems under uncertainty. I investigate those solution methods with respect to their convergence properties and am especially interested in integrating math programming (for combinatorial action spaces) and reinforcement learning. My application area of interest is healthcare logistics, especially appointment or surgery scheduling. Further, I teach the Master course reinforcement learning.
The topics and applications covered by SOR members give many directions for research for BSc and MSc students in internships and final projects. Some of these projects may be carried out in companies such as Arriva, Booking.com, DAT. mobility, KLM, NS, Ortec, Rhythm, Thales, TNO, but also in a societal environment such as a hospital.