Martin, Jan Niklas and Wilsdorf, Pia and Uhrmacher, Adelinde M. (2025) A Quantitative Metric to Evaluate the Satisfaction of Stylized Facts. In: 1st International Workshop on Behavioural Metrics and Quantitative Logics, 25 August 2025, Aarhus, Denmark.
Full text not available from this repository.Abstract
Stylized facts offer a way for validating and calibrating simulation models in data-scarce applications. SF-DSL formalizes such facts for automated model checking against simulation output. This paper extends SF-DSL with quantitative semantics to compute robustness and penalty scores, enabling optimization-based calibration. A case study using an epidemiological model demonstrates how these metrics support the identification of parameter configurations that best satisfy behavioral requirements.
| Item Type: | Conference or Workshop Item (Paper) |
|---|---|
| Additional Information: | accepted |
| Projects: | E-MoSi-CA |
