Machine learning could solve riddles of galaxy formation

A new machine-learning simulation system developed at the University of Illinois promises cosmologists an expanded suite of galaxy models - a necessary first step to developing more accurate and relevant insights into the formation of the universe.