ch Pajot's page

About me

Arthur Pajot

I'm currently pursuing a P.h.D at LIP6 under the supervision of Patrick Gallinari .

My area of interest are Machine Learning (Deep), and more generally artificial intelligence. I try to forecast times series using domain specific knowledge. I like to train models.

Recent News

  • November 2017 Our Paper, Towards a Hybrid Approach to Physical Process Modeling has been accepted at NIPS worshop for physical science.
  • November 2017 : I recently published a short vulgarisation paper about machine learning and weather forecast in La Recherche, an excellent French popular science magazine.
  • 21 September 2017 : I gave a talk about forecasting SST at the France/Japan Machine Learning Workshop at ENS Ulm,

The class I currently teach

Element de programmation

Page du cours

Apprentissage statistique (M2 Mathématiques - Certificat Big Data)

Page du cours


  • Emmanuel De Bezenac, Arthur Pajot, Patrick Gallinari (2017). Deep Learning for Physical Processes: Incorporating Prior Scientific Knowledge. Arxiv
  • Arthur Pajot, Ali Ziat, Ludovic Denoyer, Patrick Gallinari (2016). Incorporating prior knowledge in Spatio-temporal Neural Network for climatic data. ,Proceedings of the 6th International Workshop on Climate Informatics , 73-76. Climatinformatics