I am currently enrolled as a PhD student for Electricité de France & INRIA Sophia-Antipolis 📚
My subject is Federated Learning for Time Series, applied to the energy demain.
More precisely: Federated Learning (FL) aims to learn machine learning models on decentralized and private data💻. This problem is easily solved when data is identically distributed and stationary, but this assumption usually does not hold for Time Series (TS). My aim is to adapt methods from the FL and TS communities to train personalized models that generalize well across users and time periods. An application is individual time series forecasting (e.g. households’ electricity consumption).
I am also a science, video & photography enthusiast 🎥 (see links to my projects).