基于模糊RBF神经网络的单位廉政风险评价研究

廉政风险防控是反腐败工作的新举措,对风险的评价是风险管理中一个不可缺少的流程.首先运用文献研究方法,提出廉政风险评价的初步指标,再运用帕累托理论得到经过约减的指标体系.建立基于模糊RBF神经网络的廉政风险水平评价模型.并在某军区对20个单位的廉政风险水平进行评价,与传统的评价模型相比,模型能够克服评估过程中的随机性和不确定性,以及计算过程中容易陷入局部极小的问题.

Unit Corrupted Risk Evaluation Base on Fuzzy RBF Neural Network

Corrupted risk control is a new anti-corruption measure,the risk evaluation is an essential risk management processes.In this paper,proposed the preliminary corrupted risk evaluation indexes using literature research methods,and then reduced the indexes using the Pareto theory.Established the level of corrupted risk evaluated model based on fuzzy RBF neural network,and evaluated the corrupted risk level of 20 units using this model in a military area.Compared with the traditional evaluation model,this model can overcome the randomness and uncertainty in the process of evaluation,and can overcome fall into local minimum problem.