Leveraging Knowledge to Design Machine Learning Despite Limited Data

Explore a conference talk on leveraging knowledge for machine learning in data-scarce environments. Delve into the challenges of implementing decision support procedures based on machine learning methods in industrial settings with limited data availability. Discover research on transfer learning and hybrid models that incorporate knowledge from related domains or physics to create efficient models with minimal data requirements. Examine successful industrial collaborations that have employed these learning models to design machine learning solutions for small data regimes and develop powerful decision support tools. Gain insights from Mathilde Mougeot of ENSIIE & Centre Borelli/ENS Paris-Saclay on overcoming data limitations in industrial machine learning applications.