改进支持向量机的计算机网络安全态势预测方法

Computer Network Security Situation Prediction Method Based on Improved Support Vector Machine

计算机网络安全态势预测结果与实际值偏差大,为此提出基于改进支持向量机的计算机网络安全态势预测方法。构建网络流量运行参数、流量变化参数、接受攻击强度参量的支持向量机,设置安全态势最大值为1,引入转化系数,计算任意时刻安全态势值,实现网络安全态势预测。在测试结果中,设计方法的预测结果与实际值的平均绝对误差和均方差分别为0.016和0.146,偏差较小,满足实际需求。

Computer network security posture prediction results deviate from the actual value,for which a computer network security posture prediction method based on improved support vector machine is proposed.The support vector machine with network traffic operation parameters,traffic change parameters and accepted attack strength parameters is constructed,the maximum value of security posture is set to 1,and the transformation coefficient is introduced to calculate the value of security posture at any moment to achieve network security posture prediction.In the test results,the average absolute error and mean square deviation of the prediction results of the design method and the actual value are 0.016 and 0.146,respectively,which are small deviations and meet the practical requirements.