SLIDE 1
Scikit-learn
some perspectives
Lundi 17 septembre 2018 Lancement de l’initjatjve scikit-learn @ Fondatjon Inria
Development
dynamics
Lundi 17 septembre 2018 Lancement de l’initjatjve scikit-learn @ Fondatjon Inria
Scikit-learn some perspectives Lundi 17 septembre 2018 Lancement - - PowerPoint PPT Presentation
Scikit-learn some perspectives Lundi 17 septembre 2018 Lancement - - PowerPoint PPT Presentation
Lundi 17 septembre 2018 Lancement de linitjatjve scikit-learn @ Fondatjon Inria Scikit-learn some perspectives Lundi 17 septembre 2018 Lancement de linitjatjve scikit-learn @ Fondatjon Inria Development dynamics & prospects Lundi
SLIDE 2
SLIDE 3
Development
dynamics
Lundi 17 septembre 2018 Lancement de l’initjatjve scikit-learn @ Fondatjon Inria
SLIDE 4
Scikit-learn : the vision
- Democratjzing « machine learning »
Mathematjcal building blocks of AI Accessible beyond mathematjcians and computer scientjsts
- A tool also for productjon
Avoid oversimplifjcatjon : we target a technical audience Quality sofuware engineering
- Open source
SLIDE 5
10 years
v0.1, released by researchers at Inria
2011
ODSC price for best tool
2010 v0.1 2011 v0.8 2012 v0.10 2013 v0.13 2014 v0.15 2015 v0.17 2016 v0.18 2017 v0.19 2018 v0.20 kNN Clustering Random Forest Modèles linéaires Détection d’anomalie Gradient Boosting Données sparses Parallélisme Sélection de modèle Calcul distribué Données catégorielles
Open Data Science Conference
Vitesse SVM
2009
1st internatjonal sprint
2017
SLIDE 6
10 years
Monthly website traffjc
SLIDE 7
A large impact
>500 000 actjve users 12 000 academic citatjons
50 30 20
Windows Mac Linux
34 63 3
Academic Industrial Other
SLIDE 8
Community-driven development
Monthly actjve contributeurs
SLIDE 9
Community-driven development
Contributor’s actjvity Actjvity Contributor
A few very actjve people Many occasional contributors
SLIDE 10
A professional core
- Inria :
- O. Grisel
- J. du Boisberranger
- G. Lemaître (50%)
- J. van den Bossche (50%)
Actjvity Contributor
- Columbia :
- A. Mueller (50%)
stagiaires
- Sydney university :
- J. Nohman (50%)
Paid to work on the project (2018): Public research money
SLIDE 11
Scikit-learn @ fondation Inria
- Goals : stjmulate the development of scikit-learn
- Mixte governance between community and sponsors
Keep the energy and the confjdence of the community
Take strategic recommendatjons from the industry
Sponsoring Mix community-driven development with industrial interest
SLIDE 12
Development
prospects
Lundi 17 septembre 2018 Lancement de l’initjatjve scikit-learn @ Fondatjon Inria
SLIDE 13
Positioning
Databases Polyvalence Computational performance Model complexity Large- scale
SLIDE 14
Scikit-learn, new ambitions
Scaling up
- Distributed computjng
- Vendor acceleratjons
Data integratjon
- Missing data
- Categorical values
Interpretability & understanding
- Model interpretatjon
- Confjdence
?
?
?
Technical committee
More frequent versions
SLIDE 15
Thanks to our partners
SLIDE 16
Alexandre Gramfort Alexander Fabisch Alexandre Passos Andreas Mueller Arnaud Joly Brian Holt Bertrand Thirion David Cournapeau David Warde-Farley Fabian Pedregosa Gael Varoquaux Guillaume Lemaitre Gilles Louppe Jake Vanderplas Jaques Grobler Jan Hendrik Metzen Jacob Schreiber Joel Nothman Joris Van den Bossche Kyle Kastner Lars Buitjnck Loïc Estève Shiqiao Du Mathieu Blondel Manoj Kumar Noel Dawe Nelle Varoquaux Olivier Grisel Paolo Losi Peter Pretuenhofer Hanmin Qin Raghav Rajagopalan Robert Layton Ron Weiss Roman Yurchak Tom Dupré la Tour Vlad Niculae Vincent Michel Wei Li And many others