Date de l'évènement :
à 12:00, (Paris, Europe)
Lieu de l'évènement :
En ligne
Professeur Roch GIORGI - Equipe QuanTIM-SESSTIM
Intervenant :
Gaël Varoquaux
Soda, Inria Saclay
Description :
With increasing health data is collected, machine learning seems a perfect tool to improve individuals' health. I will summarize what we have learning working on medical images and electronic records. Behind our prediction successes, we have been faced with the challenges of data preparation, of censored data and other biases, and of the need for causal inference.
Key words: machine learning, health, electronic health record, censoring, biases
