Data science & artificial intelligence
For digitalized industry & services
Cette journée proposera un aperçu des recherches réalisées au sein de la chaire DSAIDIS en donnant la parole aux différents membres de l’équipe académique qui la compose. Les présentations seront suivies d’une session de questions/réponses, puis d’un temps d’échange libre pour développer avec les partenaires de la chaire les sujets abordés et les ouvrir sur des problématiques industrielles.
Les inscriptions sont réservées aux entreprises partenaires (Airbus, ENGIE, IDEMIA, Safran et Valeo) et se feront sur l’extranet de la chaire.
9h – 9h30 Welcome Coffee and Introduction
9h30 – 10h50 – Axis 2 Exploiting large scale, heterogeneous partially labeled data
• Hicham Janati: Optimal Transport for Machine Learning (50min)
• Luc Brogat-Motte: Learning to Predict Graphs with Fused Gromov-
Wasserstein Barycenters (30min)
10h50 – 11h10 Coffee Break
11h10 – 12h35 – Axis 1 Building predictive analytics on time series and data streams
• Nathan Huet: Functional Extremes or Karhunen-Lo`eve Expansion for Ex-
treme Data (25min)
• Emilia Siviero: A Statistical Learning View of Simple Kriging (25min)
• Dimitri Bouche: Wind power predictions from nowcasts to 4-hour forecasts:
a learning approach with variable selection (25min)
12h35 – 14h Lunch (room 0A202)
14h – 15h20 – Axis 4: Learning through interations with environment
• Marc Hulcelle: Computational Multimodal Models of Users’ Affective Trust
in Multiparty Human-Robot Interaction (25min)
• Aina Gari Soler: Polysemy in Spoken Conversations and Written Texts
(25min)
• Gael Guibon: A Semi-supervised Framework Dedicated to Customer Service
(25min)
15h20 – 15h40 Coffee Break
15h40 – 17h – Axis 3 Machine Learning for trusted and robust decision
• Anass Aghbalou: Tail inverse regression for dimension reduction with ex-
treme response (25min)
• Jayneel Parekh: Listen to Interpret: Post-hoc Interpretability for Audio Net-
works with NMF (25min)
• Jean-Rémy Conti / Nathan Noiry: Learning an Ethical Module for Bias
Mitigation of pre-trained Models (25min)
17h-17h15 – Discussion and conclusion.