GENOPT 2016: Design of a generalization-based challenge in global optimization

While comparing results on benchmark functions is a widely used practice to demonstrate the competitiveness of global optimization algorithms, fixed benchmarks can lead to a negative data mining process. To avoid this negative effect, the GENOPT contest benchmarks can be used which are based on randomized function generators, designed for scientific experiments, with fixed statistical characteristics but individual variation of the generated instances. The generators are available to participants for off-line tests and online tuning schemes, but the final competition is based on random seeds communicated in the last phase through a cooperative process. A brief presentation and discussion of the methods and results obtained in the framework of the GENOPT contest are given in this contribution.