EURO Doctoral Dissertation Award 2010

The prize has been conferred at the EURO 2010 Conference taking place in Lisbon (Portugal)

The EDDA 2010 Jury consisting of S. Martello, M.Rönnqvist (Chair), A. Salo, H. Stadtler, J. Teghem selected three finalist doctoral dissertation:

1 - Operating room planning and scheduling: solving a surgical case sequencing problem

Brecht Cardoen, Vlerick Leuven Gent Management School & Faculty of Business and Economics, Katholieke Universiteit Leuven, Reep l, B-9000, Gent, Belgium, brecht.cardoen@vlerick.be

We present the main results of the author's PhD dissertation, which was defendedt the Katholieke Universiteit Leuven and supervised by E. Demeuleemeester. The thesis studies the impact of planning and scheduling procedures on a hospital's operating room performance. It incorporates an extensive review of both scientific contributions and the current practice of hospitals in Flanders (Belgium). The emphasis of the research, though, is directed towards the development, the testing and the application of exact and heuristic algorithms, such as dedicated branch-and-bound procedures or mixed integer linear programming approaches, for surgery sequencing in a day-care environment

2 - Application-oriented Mixed Integer Non-Linear Programming

The main topic of the thesis is Mixed Integer Non-Linear Programming, with focus on non-convex problems (i.e. problems for which the feasible region of the continuous relaxation is a non-convex set) and real world applications. Different kinds of algorithms are presented: linearization methods, heuristic and global optimization algorithms. Also, different kinds of real-world applications are solved, arising, for example, from Hydraulic and Electrical Engineering problems. The last part of the thesis is devoted to software and tools for mixed integer non-linear programming problems.

3 - A mathematical contribution of statistical learning and continuous optimization using infinite and semi-infinite programming to computational statistics

Sureyya Ozogur-Akyuz, Department of Mathematics and Computer Science, Bahcesehir University, Bahcesehir University, Dept of Mathematics and Computer Science, Ciragan cad. Besiktas, 34353, Istanbul, Turkey, sureyya.akyuz@bahcesehir.edu.tr

ln Machine Learning algorithms, one of the crucial issues is the representation of the data. As the data become heterogeneous and large-scale, single kernel methods become insufficient to classify nonlinear data. Convex combinations of kernels were developed to classify this kind of data. Nevertheless, selection of the finite combinations of kernels is limited up to a finite choice. ln order to overcome this discrepancy, we propose a novel method of "infinite' kernel combinations by infinite and semi-infinite programming regarding ail element8 in kernel space. This provides to study variations of combinations of kernels when considering heterogeneous data in real-world applications.

"Application-oriented Mixed Integer Non-Linear Programming" by Claudia D'AMBROSIO (Italy).