"Learning the properties of our Galaxy with Stellar Streams"

The Galactic halo is criss-crossed by long stellar streams that are probably the remnants of defunct globular clusters and dwarf galaxies. I will present the recent discoveries of these structures including several streams that can be traced back to well known halo features. While streams clearly inform us in a direct way about past accretions onto our Galaxy, their most promising property is that they allow us to measure the Galactic acceleration field and they may possibly allow us to reveal the presence of small-scale of dark matter overdensities in the halo. I will present a novel unsupervised machine-learning method that we have developed that fits the acceleration field from these streams while also learning the transformation from observed kinematic coordinates to canonical action-angle variables.