Data Science Liège MEETUP #15
The mission of Data Science Liège is to offer a forum, upon which participants can leverage to federate data science initiatives, showcase projects and ideas, call for support and partnerships, disseminate knowledge and stimulate public awareness.
In order to comply with the measures taken by our government in the struggle against Covid-19, this meetup will be held online.
Participation is free but registration is required.
Follow us on Twitter : @DSLiege for updates and latest news!
PS: We are looking for participants to present and showcase their past or ongoing data science projects. Feel free to submit your propositions by email to datascience@uliege.be.
Programme
- 18:00 Introduction
- 18:05 Short talks
- 19:15 Panel discussion
- 19:30 End of the meetup
Intervenants
-
David VANDEN ABEELE
Partner and senior quantitative modeller // Credo Software -
Fabien FRANCIS
Data Scientist // Micropole Consulting Belgium
MLOps – Data Science on Steroids – David Vanden Abeele
Most data-driven organizations make money not from a few fantastic AI models, but rather from the combination of many machine learning models and scripts that run continuously and require constant maintenance. Keeping tens or hundreds of such ‘machine learning products’ in the air only works if you are organized for it and supported by appropriate tooling.
MLOps, the Data Science equivalent of DevOps, offers a promising solution to these challenges. This talk discusses the benefits and pitfalls of building and deploying machine learning products in an MLOps fashion, based on practical experiences in the financial sector.
Quels sont les nouveaux services d’AWS qui vont accélérer l’utilisation de l’IA dans votre Business et faciliter la vie des Data Scientists ? – Fabien Francis
Amazon Web Services (AWS) a été nommée en 2020, leader du « Magic Quadrant for Cloud AI Developer Service» de Gartner.
Durant le mois décembre a eu lieu « L’événement » d’AWS », re:Invent 2020.
L’offre d’AWS répond aux besoins à la fois de ceux qui n’ont pas encore de compétences en ML et de ceux qui recherchent des fonctions avancées.
Ceux qui n’ont pas encore de compétences en ML peuvent utiliser des services d’IA préentrainés avec des API qui comprennent un large éventail de fonctions de vision, NLP, Speech To text, etc. Les développeurs et Data Scientists à la recherche de fonctions plus avancées peuvent utiliser la suite de ML complète de SageMaker.
Les annonces de nouveaux services de Machine learning ont fusé de toute part, et nous essayerons d’en ressortir l’essentiel pour vous.