Risk warning model based on radial basis function neural network

General contracted project is a very complicated system. The risk of general contractor gets greater and greater with the technology and free trade pressure after China enters WTO, so how to escape and disperse risk becomes a hot topic. Risk warning model is valuable to risk analysis because it can make general contractors have risk management capacity. A new method to analyze the risk of general contracted project is proposed, which uses the radial basis function (RBF) neural network to establish a risk warning model. The analytical hierarchy process (AHP) is carried out to determine the key factors, which are used as inputs of radial basis function neural network through numeralization. By training the neural network, we obtain the nonlinear function from key factors to risk compensation, and then the sensitivity of key factors is launched. The practical project sample analysis proves the validity of the method.