基于Stacking集成学习的网络安全态势预测方法

针对现有的网络安全态势预测模型预测精确度低且泛化能力差等问题，提出一种基于Stacking模型融合的态势预测方法。该方法中，借助Stacking算法将TCN网络、WaveNet、GRU、LSTM进行集成挖掘态势数据之间的相关性；之后利用逻辑回归进行预测得到最终态势值；利用粒子群优化算法进行参数寻优，提升模型性能。基于2个数据集进行验证，实验表明，所提预测方法具有较小的均方误差和平均绝对误差，收敛速度较快，拟合度均可达0.999，可以很好解决预测精确度低的问题，提升了模型的泛化能力。

Research on Network Security Situation Prediction Method Based on Stacking Integrated Learning

To address the problem of low prediction accuracy of existing network security posture prediction models, a prediction method based on Stacking model fusion is proposed. In this method, the TCN network, WaveNet, GRU, and LSTM are integrated with the Stacking algorithm to explore the correlation among the situational data; after that, logistic regression is used to further predict the final situational values; the particle swarm optimization algorithm is used to optimize the parameters and improve the model performance. Based on two data sets for validation, the experiments show that the proposed prediction method has small mean square error and mean absolute error, fast convergence speed, and the fit degree can reach 0.999, which can well solve the problem of low prediction accuracy.