Applying DevOps to Machine Learning: ROCKFlows, a Story from the Trenches

Abstract : The Machine Learning (ML) community is currently blooming with hundreds of new algorithms to implement tasks such as data classification for example. To support data scientists and engineers who have to chose among all these algorithms, we are defining the ROCKFlows platform to automatically create a software product line of workflows integrating such algorithms. This paper describes the approach that support this approach in a devops way.
Document type :
Other publications
Liste complète des métadonnées

Cited literature [3 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01792775
Contributor : Sébastien Mosser <>
Submitted on : Tuesday, May 15, 2018 - 6:02:48 PM
Last modification on : Monday, November 5, 2018 - 3:52:09 PM
Document(s) archivé(s) le : Tuesday, September 25, 2018 - 3:56:30 AM

File

applying-devops-machine.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01792775, version 1

Collections

Citation

Mireille Blay-Fornarino, Günther Jungbluth, Sébastien Mosser. Applying DevOps to Machine Learning: ROCKFlows, a Story from the Trenches. 2018. ⟨hal-01792775⟩

Share

Metrics

Record views

104

Files downloads

122