MASCOT: A Multiagent System for Hybrid Optimization
Sven Hader · 1997
In this paper, the so-called meta optimization problem, that is, the task of choosing a suitable optimization technique as well as favorable strategy parameters for an optimization problem at hand, is described. It is stated that under conditions most often found in practical optimizations hybrid techniques are best suited. The optimization system MASCOT is introduced that uses a hybrid technique adapting dynamically to concrete problems. For illustration, the optimization of a multi-location inventory model using MASCOT is shown.