Abstract: In order to obtain a high-quality schedule, an intelligent personnel scheduling system combining deep neural network and branch and bound method is designed. The characteristics and difficulties of personnel scheduling problem are introduced; the integer programming model of the problem is constructed; a deep neural network-assisted branch and bound method is proposed. It can make reasonable branch selection and branch pruning at each step in branch and bound by learning the existing personnel scheduling problem with known optimal solutions. The proposed method is an innovative attempt to solve combinatorial optimization problems with deep learning method. The experimental part verifies the feasibility of this method on different parameter settings and standard instance sets.