Hybrid model by RS_RBF evaluate the investment risks of High-tech projects
Lianghai Chen · 2011
In order to resolve the redundant information in High-tech project evaluation, a high-tech investment risk evaluation model combining a rough sets RS and the RBF neural network is presented. First using the rough set's powerful numerical analysis capabilities, this model does the attribute reduction on the evaluation index which reduces the training data of RBF neural network and simplifies the network structure. Then this model trains the data after reduction using the RBF neural network. Last applying this model to the High-tech project evaluation, the simulation results show that compared with the RBF neural network model, the hybrid model can achieve more satisfactory results such as speeding up the network operator speed, minimizing the evaluation error, and improving the evaluation precision.