Research on Risk Evaluation Method of Discrete Uncertainty Attribute Based on Rough Set and Neural Network in Construction Project

摘要

In large-scale project management, the risk management has discrete uncertainty attribute. In order to deal with the risk management and complex decision making caused by risk factors and uncertainty of construction project, using the theory and method of rough set and neural network, its risk index system based on the balanced scorecard was reduced, the knowledge discovery and data mining technology were used to establish a risk evaluation model of construction project. According to the experimental data of construction project, the reduced index was input in neural network for intelligent training. The risk evaluation sample was input in the trained network. The final risk evaluation value of construction project can be exported. The project contractor can be selected for decision making. Through the empirical analysis, the effectiveness has been verified, and the issues of decision-making for discrete uncertainty data in large-scale project management have been resolved.