A machine learning approach to the simulation of inter-city corporate networks in mainland China

These data are used to implement the random forest algorithm to simulate the inter-city corporate networks created by Fortune China 500 firms in Mainland China. The connectivities in the networks are computed using the interlocking network model and treated as the target variables. City factors and geographical factors are treated as features. The model was trained using the 2010 training set, and subsequently verified using 2010 and 2017 test sets.