The decentralization, diversity, dynamic and combination of the outsourcing resources in the current complex outsourcing environment put forward higher requirements for the management of the manufacturing enterprises, which makes the traditional methods difficult to meet the needs of outsourcing supplier evaluation in the current situation. In this paper, a decision making model for outsourcing supplier evaluation based on hybrid PSO-Adam neural networks is proposed to solve the defect of multiattribute decision making methods, mathematical programming methods and classical neural networks methods. In order to validate the validity of this model, an example of the building materials manufacturing enterprises outsourcing supplier evaluation is carried out and compared with other existing algorithms.The results show that this model can solve the problem of evaluation of suppliers objectively and efficiently in the current complex outsourcing environment, and reduce the reliance on personal experience, the difficulty of supplier evaluation and the cost of supply chain management.