摘要
针对电子政务信息系统风险评估的复杂性和不确定性,作者提出了小波神经网络评估方法,建立了小波神经网络评估模型,最后采用Matlab7.1工具进行了仿真实验,并分别从网络的收敛速度、训练精度和训练效果上与BP神经网络算法作了对比。结果证实该方法有效克服了传统方法中人为分配权重带来的主观性,使评价结果更科学,更合理。
This paper analyzes the complexity and uncertainty of the risk factors in information system, presents a Wavelet Neural Network（WNN） method which can be used to the security risk evaluation of information system, creates a corresponding model and has done simulated experiments by Matlab7.1. Besides, the author makes a comparison between WNN and BP algorithm respectively in convergent speed, training precision and effect. The study of the case shows that, with this method, subjective judgment in previous research is avoided and the results are more scientific and reasonable.
出处
《石家庄铁道学院学报》
2007年第3期69-73,共5页
Journal of Shijiazhuang Railway Institute
基金
河北省自然科学基金资助项目(03579325A)
