Abstract: With the development and progress of Internet theory and technology, security has become an extremely important factor in the study of cyberspace. Network security situation prediction can integrate different levels of security factors and directly reflect the overall situation of network security. In order to increase the accuracy of the prediction of network security situation, this paper proposes a method to apply the deep self-encoder neural network algorithm to the network security situation prediction technology, and uses the National Internet Emergency Response Center security data set to perform simulation experiments on the matlab platform. The results show that the network security situation prediction model based on depth self-encoder has a fast learning speed and high accuracy ,which can predict the network security situation well.

Key words: automatic encoder, depth automatic encoder, network security, situation prediction, the neural network

摘要： 随着互联网理论与技术的发展与进步，安全已经成为网络空间研究中极端重要的因素.网络安全态势预测可以整合不同层次的安全因素，直观地反映网络安全的整体情况.为了增加预测网络安全态势的精度，提出了一种将深度自编码器神经网络算法应用到网络安全态势预测技术中的方法，利用国家互联网应急响应中心的安全数据集，在matlab平台上进行仿真实验.结果表明，基于深度自编码器的网络安全态势预测模型具有较快的学习速度和高准确率，可以很好地预测网络安全状况.

关键词: 自编码器, 深度自编码器, 网络安全, 态势预测, 神经网络