Astronomy and cosmology have firmly entered the data rich regime. We measure photometric properties of more than hundreds of millions of galaxies and stars. Machine learning is a data mining approach to make sense of this data flood, and has found many applications in cosmology, from the measurements of photometric redshifts, to the distinguishing between stars and galaxies. In the Physical Cosmosolgy group we have been pushing the boundaries of both the applications of ML in cosmology, and also devising new ML algorithms. We have strong connections with other research departments at the LMU, and have produced students well equipped for a job outside of acedemia.