1.8 KiB
1.8 KiB
tags
| tags | ||
|---|---|---|
|
(17) (PDF) Machine Learning Pipelines: Provenance, Reproducibility and FAIR Data Principles authors: year: doi: zotero: (Open) URL: https://www.researchgate.net/publication/342377391_Machine_Learning_Pipelines_Provenance_Reproducibility_and_FAIR_Data_Principles abstract: Machine learning (ML) is an increasingly important scientific tool supporting decision making and knowledge generation in numerous fields. With this, it also becomes more and more important that the results of ML experiments are reproducible. Unfortunately, that often is not the case. Rather, ML, similar to many other disciplines, faces a reproducibility crisis. In this paper, we describe our goals and initial steps in supporting the end-to-end reproducibility of ML pipelines. We investigate which factors beyond the availability of source code and datasets influence reproducibility of ML experiments. We propose ways to apply FAIR data practices to ML workflows. We present our preliminary results on the role of our tool, ProvBook, in capturing and comparing provenance of ML experiments and their reproducibility using Jupyter Notebooks.
Here the focus is about reproducibility of ML experiments.
- They propose ProvBook to enable ML experiment via Jupyter Notebooks.
They aim at applying FAIR principle to ML workflows
Why reproducibility of ML experiments is important?
Because more and more decision making and knowledge extraction is based on ML. Only reproducible ML experiments are trustworthy.
There is a Reproducibility crisis of ML experiments
Concerning reproducibility, the main issue is that building ML pipelines requires constant adjustments in algorithms, models, and parameter tuning.