基于机器学习和仿真优化的信贷工厂人员调度系统

systematic solution based on machine learning and simulation optimization is proposed to solve the staff allocation problem in credit factory. Firstly, the credit factory is modeled as a combination of multiple queues, and a credit factory simulation system is built based on queuing theory and discrete simulation theory. Secondly, we use the prophet model to predict the credit order arrival rate of the next period of and input it into the simulation system. Thirdly, based on Bayesian optimization algorithm, we optimize the worker allocation problem of the simulation system. Finally, from the simulation system, we obtain the optimized worker allocation scheme in the next period, and then apply this scheme to the real credit factory. The experimental results show that the approval efficiency of the optimized algorithm is 166.7% and 38.8% higher than that of random setting and human experience adjustment, respectively.