基于CS-RBF神经网络的网络安全态势预测方法研究

Research on Network Security Situation Prediction Method Based on CS-RBF Neural Network

为了提高网络安全态势预测精度,帮助企业组织保护自身隐私数据,研究提出将布谷鸟算法与径向基神经网络算法融合设计出网络安全态势预测算法,并采用三角模糊数计算指标权重,设计出基于模糊层次分析法的网络安全态势感知指标体系,用于给预测模型提供输入数据。仿真实验结果显示,此次研究设计的网络安全态势预测模型在训练收敛后的损失函数值为0.82,对于对比模型,且在测试集上的平均绝对误差值与绝对误差标准差分别为1.82与4.57,也明显低于对比模型。此次研究设计出的网络安全态势预测模型具有良好的预测精度,对于提升我国企业数据安全水平具有应用潜力。

In order to improve the accuracy of network security situation prediction and help enterprise organizations protect their private data,the research proposes to design a network security situation prediction algorithm by integrating the cuckoo algorithm and the radial basis function neural network algorithm,and calculates the index weight using triangular fuzzy numbers.A network security situation awareness index system based on the fuzzy analytic hierarchy process is designed to provide input data to the prediction model.The simulation experiment results show that the loss function value of the network security situation prediction model designed in this study after training convergence is 0.82.For the comparison model,the average absolute error value and absolute error standard deviation on the test set are 1.82 and 4.57,respectively,which are also significantly lower than the comparison model.The network security situation prediction model designed in this study has good prediction accuracy,and has application potential for improving the data security level of Chinese enterprises.