“…From the abovementioned MGI database of Sm-Co materials, a data set of saturation magnetization was extracted (which is provided in Section S1 in the Supporting Information), and it contains data of 214 kinds of materials from various resources published in 40 years from 1978 to 2018. ,,,,− The information related to saturation magnetization includes the Sm-Co matrix with a certain atomic ratio ( x matrix = ( x Co + x M )/ x Sm ), doping element (denoted as “M”), doping concentration ( x doping = x M / x Co ), processing of material ( C proc ), form of the material for testing ( C form ), grain size ( d ), and so on. It is known that, for a data set, machine learning can generate a predictive model, which can be applied to design new materials.…”