Abstract: Because of the lack of risk assessment tools for project contractors, often loss occurs in such companies. Some time, the loss may be great. To solve this problem, a risk assessment system for project contractors is built by analyzing their characteristics. By this system, the risk assessment is done by using generalized regression neural network (GRNN). By using Matlab as tool, a case from a company of building security is given to demonstrate the effectiveness of the proposed approach. 

Key words: risk assessment, generalized regression neural network (GRNN), matlab