Description
In this work we introduce Reskit (researcher’s kit), a library for creating and curating reproducible pipelines for scientific machine learning. A natural extension of the Scikit Pipelines to general classes of pipelines, Reskit allows for the efficient and transparent optimization of each pipeline step. Its main features include data caching, compatibility with most of the scikit-learn objects, optimization constraints such as forbidden combinations, and table generation for quality metrics. Reskit’s design will be especially useful for researchers requiring pipeline versioning and reproducibility, while running large volumes of experiments.