本文对2000-2015年徐州社会经济发展主要指标及其数据进行研究,采用BP神经网络构建了自助预测模型,并在此基础上,改进了神经网络在神经元连接过程中样本数据自相关的问题,消除了数据趋势性特征。改进方法主要以灰色关联法为切入点,重点考察了MS与GDP、ZG、PG之间的协同性。研究结果表明,目前徐州经济发展与职业教育协同度较低,整体处在工业经济向知识经济过渡时期,职业教育规模对明瑟收益率贡献低于非职业教育。 This paper studies the main indicators and their data of social and economic development in Xuzhou from 2000 to 2015, constructs a self-service prediction model based on BP neural network, and on this basis, improves the problem of auto-correlation of sample data in neuronal connection process , Eliminating the data trend characteristics. The improvement method mainly takes the gray relational method as the starting point, and focuses on the synergy between MS and GDP, ZG and PG. The results show that the current low degree of synergy between economic development and vocational education in Xuzhou, the overall in the transition from industrial economy to knowledge-based economy, vocational education scale contribution to the rate of return of Moses than non-vocational education.