Study on Risk Evaluation of High-tech Projects Investment Based on RS_RBF
Chen Liang-hai · Jisuanji fangzhen · 2010
In order to resolve the redundant information in High-tech project evaluation,rough sets based RBF neural network nodel are presented for the evaluation model of high-tech investment risk..With the powerful numerical analysis capabilities of rough set,the attributes of evaluation indexes are reduced,which decreases the training data of RBF neural network,thereby simplifies the structure of RBF neural network and speeds up the training speed.Simulation results show that the hybrid model can achieve more satisfactory results,and compaired with the standard RBF neural network model,the hybrid model has obvious advantages.