Reconstructing the three-dimensional structure of star-forming cores with the help of neural networks A novel machine learning approach shows promise in helping astronomers infer the internal structure of stellar nurseries from telescope observations and to gather new insights into star formation. In galaxies, stars are born inside enormous clouds of interstellar gas and dust that are called molecular clouds. Through turbulence and gravitational instabilities, these molecular clouds develop intricate substructures, in which the formation sites of stars are located. The densest and smallest of these structures are the star-forming cores, the direct progenitors of individual stars or entire star clusters. Inferring the structure of these clouds and cores is a key aspect in understanding star formation.
Ksoll et al. (Thu,) studied this question.