针对传统网络安全态势感知预测方法预测精度低的问题，本文提出了一种基于PSO-LSTM的网络安全态势预测方法。网络安全态势数据由于其具有时序性，过往的态势值往往影响到未来的网络状况。LSTM网络可以有效地处理具有时序性特征的数据，同时依靠PSO算法优秀的全局搜索能力优化LSTM网络的超参数。仿真实验表明，本文提出的方法有效地提高了态势值预测的准确度。 Aiming at the low accuracy of the traditional network security situation perception prediction method, this paper presents a network security situation prediction method based on PSO-LSTM. Due to the timing of network security situation data, the past situation values often affect the future network situation. LSTM network can effectively process data with temporal characteristics, and at the same time, the super parameters of LSTM network can be optimized by the excellent global search capability of PSO algorithm. Results show that the method proposed in this paper can effectively improve the accuracy of situation value prediction.
展开▼
机译：针对传统网络安全态势感知预测方法预测精度低的问题，本文提出了一种基于PSO-LSTM的网络安全态势预测方法。网络安全态势数据由于其具有时序性，过往的态势值往往影响到未来的网络状况。LSTM网络可以有效地处理具有时序性特征的数据，同时依靠PSO算法优秀的全局搜索能力优化LSTM网络的超参数。仿真实验表明，本文提出的方法有效地提高了态势值预测的准确度。 Aiming at the low accuracy of the traditional network security situation perception prediction method, this paper presents a network security situation prediction method based on PSO-LSTM. Due to the timing of network security situation data, the past situation values often affect the future network situation. LSTM network can effectively process data with temporal characteristics, and at the same time, the super parameters of LSTM network can be optimized by the excellent global search capability of PSO algorithm. Results show that the method proposed in this paper can effectively improve the accuracy of situation value prediction.
展开▼
