Prof. Bartz-Beielstein: External Examiner at Warwick Business School

wbs-logoProf. Thomas Bartz-Beielstein was invited to Warwick Business School (WBS) last Monday. WBS is part of the University of Warwick, one of the UK’s top universities. Thomas served as an external examiner for Jawal Elomaris’s thesis “Efficient Learning Methods to Tune Algorithm Parameters”. The examination was for the “Doctor of Philosophy in Warwick Business School”.  Dr. Bo Chen Professor of Operational Research and Management Science, served as the internal examiner and Dr. Victor Podinovski, Professor of Operational Research, was appointed as an examination advisor. The thesis was supervised by Dr. Jürgen Branke, Professor of Operational Research & Systems.

wbs1

Prof. V. Podinovski, J. Elomari, Prof. T. Bartz-Beielstein, and Prof. B. Chen

The thesis focuses on the algorithm configuration problem. Incorrectly setting algorithm parameters may cause the algorithm not to work effectively and efficiently. Several approaches for online and offline tuning were discussed in this thesis. Methods for improving racing algorithms [1] were developed. These, namely Optimal Computing Budget Allocation (OCBA), methods are are also relevant for SPOT. OCBA enhances simulation and optimization procedures efficiency by optimally allocating the budget [2].

OCBA is a popular tool for handling noise in simulation and optimization settings. In 2007, Christian Lasarczyk presented the first approach, which integrates OCBA into SPOT. Two years earlier, Thomas Bartz-Beielstein, Daniel Blum, and Jürgen Branke combined OCBA and particle swarm optimization [4].

References

[1] M. Birattari, T. Stu ̈tzle, L. Paquete, and K. Varrentrapp. A racing algorithm for configuring metaheuris- tics. In W. Langdon, editor, GECCO 2002: Proceedings of the Genetic and Evolutionary Computation Conference, pages 11–18, San Francisco CA, 2002. Morgan Kaufmann.

[2] C. H. Chen. Stochastic simulation optimization: an optimal computing budget allocation. World Scientific, 2010.

[3]  C. W. G. Lasarczyk. Genetische Programmierung einer algorithmischen Chemie. PhD thesis, Technische Universität Dortmund, 2007.

[4] T. Bartz-Beielstein, D. Blum, and J. Branke. Particle swarm optimization and sequential sampling in noisy environments. In R. Hartl and K. Doerner, editors, Proceedings 6th Metaheuristics International Conference (MIC2005), pages 89–94, Vienna, Austria, 2005.