Regression Testbed for Reproducible Research

Highlights

• standardized interfaces for datasets (including meta-data) and algorithms
• management of generated artifacts and configuration/parametrization options for full data/model lineage and traceability
• utility tools for analyzing and visualizing results along any experiment dimension (dataset variants, algorithm variants, parametrization variants, etc.)

---

Technology Stack: Python, HDF5, Pandas, SLURM, Dask Distributed, Scikit-Learn

Work Affiliation: German Aerospace Center (DLR)