基于SVM的网络威胁频率预测算法研究

Research on SVM-Based Algorithms for Network Threat Frequency Prediction

网络安全风险评估分为定量评估与定性评估两种方法。现今的评估方法多数仅对网络安全事件进行定性的评估,不能给出网络安全风险值的定量说明,评估结果缺乏客观性、动态性和可信性。本文在深入研究风险评估理论和实验的基础上提出了一种新的网络安全定量风险评估方法——基于支持向量机SVM的网络威胁频率预测方法,构造了相应的SVM模型,并采用仿真数据对模型进行了验证。该模型对来自某局域网IDS实测数据的网络威胁频率进行了预测,和真实值的对比说明,所提出的网络威胁预测算法是行之有效的。

There are two types of risk assessment methods： the quantitative and qualitative ones. Most risk assessment methods only give a qualitative account of the network threat events without the quantitative value, and the evaluation resuits lack objectivity, dynamics and credibility. This paper presents a new quantitative assessment method, which is a predicting algorithm for network threat frequency based on SVM. A model of support vector regression machine is constructed, and the performance of the model is valida：ed based on simulation data. The SVM model predicts the threat frequency using a LAN＇s detected data by IDS, compared with the detected data, the result shows that our method is feasible.