良好的信用环境对于促进建筑业的持续健康发展至关重要。在政府监管视角下,分析建设市场主体信用评价要点并从五个方面构建了信用评价指标体系。在此基础上,基于BP神经网络构建了建筑市场主体信用评价模型。选取国内30家建筑企业信用数据对模型进行验证,利用构建的BP神经网络模型并进行了训练和测试。分析测试结果显示,误差数据满足目标要求,基于BP神经网络的信用评价模型具有较好可行性与准确性。

A good credit environment is crucial to the sustainable and healthy development of the construction industry. From the perspective of government supervision, analyzes the main points of credit evaluation for construction market players and constructs a credit evaluation index system from five aspects. On this basis, a credit evaluation model for construction market players was constructed based on BP neural network. The 30 domestic construction companies" credit data were selected to verify the model, and the BP neural network model constructed was used for training and testing. Analysis and test results show that the error data meets the target requirements, and the credit evaluation model based on BP neural network has good feasibility and accuracy.

F426，F281，C935

国家社科基金“工程建设市场主体社会化监管机制研究”（17BGL156）；国家社会科学基金青年项目“市场决定性作用下PPP项目协同监管模式研究”（15CJL023）；广东省水利科技计划创新项目“广东水利工程建设事中事后监管模式研究”（2017-04）；贵州省水利科技计划项目“全生命周期下贵州省水利工程管理智能化建设内容与关键技术研究”（KT201601）；中央高校基本科研业务费专项资金资助“行政审批制度下工程建设政府监管机制创新”（2017B35214）。