School of
Information Technology and Electrical Engineering

Speaker: Prof. Peter Bosman
Seminar Date: Fri, 02/11/2018 - 11:00
Venue: 49-502 - AEB Seminar Room
Host: A/Prof Marcus Gallagher

Seminar Type: Guest Research Seminar


In this talk Prof. Bosman will introduce the Gene-pool Optimal Mixing Evolutionary Algorithm (GOMEA) and illustrate its advantageous properties over classical "blind" Evolutionary Algorithms (EAs), showing that under certain conditions GOMEA can scale to solve problems with millions of variables in less than an hour on a normal desktop computer.

GOMEA belongs to the class of Model-Based EAs (MBEAs) and focuses particularly on efficiently learning and exploiting so-called linkage models that describe dependencies between the variables that are used to encode solutions to the optimization problem at hand.

Recent work expands the GOMEA family as originally introduced for classical binary representations to other domains, including permutations, real values, and tree-based genetic programming for both single- and multi-objective optimization problems.

Besides the foundations of GOMEA, Prof. Bosman will present the projects that my subgroup of Medical Informatics at CWI is currently involved in, outlining how our GOMEA research also fuels our research into real-world-applied optimization, machine learning, and explainable AI in the medical domain, in collaboration with hospitals and other academic partners. In particular, I will show how we can obtain better radiotherapy treatment plans for treating cancer faster, and more insightfully with GOMEA than is currently possible in clinical practice.


Prof.dr. Peter A.N. Bosman is a senior researcher heading the Medical Informatics (MI) subgroup of the Life Sciences and Health (LSH) research group at the Centrum Wiskunde & Informatica (CWI) (Center for Mathematics and Computer Science) located in Amsterdam, the Netherlands.

He further has a part-time professor (in Dutch: deeltijdhoogleraar) position at Delft University of Technology in the Algorithmics group of the Department of Software Technology in the Faculty of Electrical Engineering, Mathematics, and Computer Science. Prof.dr. Bosman was formerly affiliated with the Intelligent Systems research group of CWI and before that with Utrecht University, where he also obtained his M.Sc. and Ph.D. degrees in Computer Science.

Prof.dr. Bosman's fundamental research focus is on the design and application of evolutionary algorithms (EAs) for single- and multi-objective optimization, and machine learning. The optimization problems considered are typically complex to an extent where a black-box optimization (BBO), or at least a grey-box optimization (GBO), perspective is required, i.e. virtually no information (BBO) or limited information (GBO) is available (or properly understood) about the optimization problem at hand. The designed EAs are moreover mostly model-based, meaning that a specific model is used to capture and exploit problem-specific features to guide the search for high-quality solutions more effectively and efficiently and get the most out of previously performed evaluations. Such models may be derived by hand or, if this isn't possible (as in e.g. the BBO case), be learned online, i.e. during optimization, using techniques from fields such as machine learning and data mining. For problems where efficient

(problem-specific) heuristic optimization techniques (i.e. local search (LS) techniques) are available or can be derived, model-based EAs are furthermore a very solid basis for hybridization to obtain the best of both worlds in terms of efficiency and effectiveness, resulting in state-of-the-art optimization algorithms for specific problems.

Prof.dr. Bosman's applied research focus is on the use of (model-based) EAs to solve key problems in the Life-Science and Health (LSH) domain that require optimization and/or machine learning. A specific focus is on improving mathematics and computer science related aspects in radiation oncology, such as automated treatment planning, deformable image registration and 3D dose reconstruction. Previous application areas have included (smart) energy and logistics, dynamic pricing of goods for revenue management, optimization of patient flows in hospitals and dynamic routing of vehicles for transportation purposes.

Prof.dr. Bosman has (co-)authored over 100 refereed publications, out of which 4 received best paper awards. According to Google Scholar, his h-index is 30 with a total of 2948 citations to his works (as measured on September 14, 2018). He is program committee member of various major conferences and journals in the EA field and related fields. In 2017, Prof.dr. Bosman was the General Chair of the main conference in the field of EAs: the Genetic and Evolutionary Computation Conference (GECCO). He has furthermore organized various workshops and tutorials on various EA related topics and has been (co-)track chair and (co-)local chair at GECCO.

Finally, the (co-)acquired research grant funding by Prof.dr. Bosman totals over €4M, which includes one STW-KWF partnership project, four NWO projects, one KiKa project, and one EIT ICT Labs project that together fund(ed) 5 postdocs, 13 Ph.D. students, a radiation therapy technologist, a scientific programmer, and various high-performance computing hardware.