Parallel and Distribution Simulation Systems
Parallel and Distribution Simulation Systems
ANSS '07 Proceedings of the 40th Annual Simulation Symposium
ParamILS: an automatic algorithm configuration framework
Journal of Artificial Intelligence Research
Automating the runtime performance evaluation of simulation algorithms
Winter Simulation Conference
| Hi-index | 0.00 |
Simulation algorithms often expose various numerical parameters, e.g., to control the size of auxiliary data structures or to configure certain heuristics. While this allows to fine-tune a simulator to a given model, it also makes simulator configuration more complex. For example, determining suitable default parameters from a multi-dimensional parameter space is challenging, as these parameters shall work well on a broad range of models. Instead of manually selecting parameter values, the configuration space of a simulation algorithm can also be searched automatically. We investigate how well ParamILS (Hutter et al. 2009), an iterated local search algorithm for algorithm configuration, can be applied to simulation algorithms, and discuss its implementation in context of the open-source modeling and simulation framework JAMES II.