作者:Benjamin Nachman et.al.
论文链接:http://arxiv.org/abs/2507.11528
发布日期:2025-07-15
解读时间:2025-07-19 18:54:42
The development of scientific data analyses is a resource-intensive process that often yields results with untapped potential for reuse and reinterpretation. In many cases, a developed analysis can be used to measure more than it was designed for, by changing its input data or parametrization. Existing reinterpretation frameworks, such as RECAST, enable analysis reinterpretation by preserving the analysis implementation to allow for changes of particular parts of the input data. We introduce FlexCAST, which generalizes this concept by preserving the analysis design itself, supporting changes to the entire input data and analysis parametrization. FlexCAST is based on three core principles: modularity, validity, and robustness. Modularity enables a change of the input data and parametrization, while validity ensures that the obtained results remain meaningful, and robustness ensures that as many configurations as possible yield meaningful results. While not being limited to data-driven machine learning techniques, FlexCAST is particularly valuable for the reinterpretation of analyses in this context, where changes in the input data can significantly impact the parametrization of the analysis. Using a state-of-the-art anomaly detection analysis on LHC-like data, we showcase FlexCAST's core principles and demonstrate how it can expand the reach of scientific data analysis through flexible reuse and reinterpretation.