Based on neural networks, the present paper gives an engineering application of data mining. The back propagation (BP) neural network is used as the algorithm of data mining. Then the effects of structural technologic parameters on stress in the weld region of the shield engine rotor in a submarine are analyzed. The mined data come from the numerical simulations of the finite element method. The effects of different parameters on the stress in the weld region are achieved from the results of the data mining. The discovered knowledge is beneficial to the security improvement of structural strength design for the engine rotor.