Machine Learning applied to Cosmology

Recent progresses in Machine Learning have unlocked new possibilities to tackle scientific problems by means of Machine Learning methods, and already many applications have been developed both in astrophysics and cosmology. In this presentation, after an introduction to Machine Learning and to the Cosmological context, I will focus on two different kinds of methods that have been applied to 1) the partitioning of the cosmic web - the dark matter density at late time - into nodes (very dense clusters of galaxies) and filaments, using probabilistic clustering technics; 2) the use of generative deep neural network in order to estimate the empirical density probability distribution. Furthermore, I will show how this last generative model can be further use for predictive purpose on the dark matter evolution: given a snapshot in the past, having z>0, it can produce the time-evolution.