Towards accurate and interpretable modelling frameworks for (astro)physics

 In this talk I will briefly present our team’s ongoing efforts to detect
and characterize the Milky Way's stellar streams, aiming to constrain
the properties of the dark matter and to unveil the ancient accretions
that they testify to. Using these stellar streams as case studies, I
will concentrate on presenting our recent efforts to devise explicable,
interpretable and accurate methods to model general physical and
astrophysical systems using modern machine learning techniques.