基于BP神经网络的网络安全评价方法研究  

Researches on the Evaluation Methods of Network Security Based on BP Neural Network

作　　者：黄丽民[1] 徐红[1] 

机构地区：[1]山东商业职业技术学院,山东济南250103

出　　处：《山东商业职业技术学院学报》2010年第4期81-83,94,共4页Journal of Shandong Institute of Commerce and Technology

摘　　要：阐述了BP神经网络的基本理论、算法、缺陷及改进方法,在遵循BP网络建模基本原则和步骤的情况下,建立了可靠、有效的网络安全综合评价模型。通过实例对网络安全进行研究,选择合适的学习样本及参数进行训练,并获得较好的评价结果,评价结果客观、正确,计算结果贴近成功案例结果。First, this paper introduced the basic theories, algorithms and detects of BP Neural Network. Complying with the basic principles and steps for establishing BP -model, it established reliable and effective network security comprehensive evaluation model. Through network security case studying, appropriate training set data and parameters selection, better evaluation results have been achieved. The evaluation results are impersonal and correct, pressing close to the results of the successful cases.

关 键 词：BP神经网络 网络安全 模型 评价指标 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP393.08[自动化与计算机技术—控制科学与工程]