ML-Based Reproducibility at the Edge

Reproducibility is a cornerstone of the scientific method. It is also a core element of compliance requirements for sensitive equipment, e.g., audit trails for medical equipment. Often, a prerequisite for computational reproducibility is the availability of software and data. However, this is problematic for edge devices whose goal is to reduce the amount of data transferred to the backend. On edge devices, by definition, it is not possible to archive all processed data. Reproducibility must thus rely on models of the data, not on the data itself.

The first step in the project is to survey existing work in areas where models are developed to replace unavailable data (e.g., machine learning, data sketching, data postdiction). The second step is to develop a ML-based solution for edge-based reproducibility in the context of a case study.