SimEx: The Art of Systematic Exploration
| Type of publication: | Inproceedings |
| Citation: | |
| Publication status: | Accepted |
| Year: | 2024 |
| Month: | September |
| Location: | PuK Workshop, Würzburg |
| Abstract: | Experimentation is essential in most research domains, serving as a good way to explore, test, and validate a system’s behaviors. However, real-world experiments typically involve physical testing, observation, or interaction with the systems, which can be dangerous, unethical, very costly, or even impossible in some cases. Therefore, in-silicon experimentation (known as computer simulations) becomes beneficial and, in most cases, cost-effective, allowing the simulation of even impossible scenarios. However, the exploration of the whole environment and all possible scenarios often becomes computationally costly and trade-offs between accuracy and computational run time has to be made. This paper proposes a systematic exploration tool that approximates the system’s behavior with the aim of significantly improving the efficiency of simulation runs in simulation studies. This tool holds the potential to revolutionize the way we approach simulation studies, making them more efficient and productive. The approximation is done using partial functions, which have been approximated based on obtained data points, i.e. results of simulation runs. We have presented an approach that allows for obtaining sufficiently good enough functional approximations while keeping the overall number of simulation runs as low as possible. |
| Keywords: | Function approximation, Structured simulations, Systematic exploration |
| Authors | |
| Added by: | [] |
| Total mark: | 0 |
|
Attachments
|
|
|
|
|
Notes
|
|
|
|
|
|
Topics
|
|
|
|
|
