中文摘要：

针对被动防御的问题，提出了一种基于支持向量回归算法的网络安全态势预测方法。在最近固定时间段的数据集数据进行归一化处理的基础上，构造样本数据并进行训练，确定预测模型的各项参数。实验证明该方法能有效地对网络未来的安全态势进行预测，结果优于其他算法，达到了实时在线条件下准确预测网络安全态势结果的目的。

英文摘要：

Aiming at the problem of passive defense, this paper presented a prediction technique in network security situation based on support vector regression algorithm. By processing normalized data set in last fixed time period, trained sample data to determine the parameter of prediction model. The experiment result shows that the technique can predict the network securi- ty situation efficiently and accurately, the result is better than other algorithms. It achieved the objective of predict network se- curity, situation accurately in real-time and on-line system.