ID	原文	译文
3093	该方法考虑异构节点的训练质量和效率，筛选恶意节点，在提升联邦学习模型准确率的同时，优化训练时延。	It considered training quality and efficien- cy of heterogeneous terminal devices, and filtrate malicious nodes to guarantee higher model accuracy and shorter train- ing delay of federated learning.
3094	首先，根据联邦学习中模型分布式训练的特点，构建基于深度强化学习的节点选择系统模型。	Firstly, according to characteristics of model distributed training in federated learning, a node selection system model based on deep reinforcement learning was constructed.
3095	其次，考虑设备训练时延、模型传输时延和准确率等因素，提出面向节点选择的准确率最优化问题模型。	Secondly, considering such factors as device training delay, model transmission delay and accuracy, an optimization model of accuracy for node selection was proposed.
3096	然后，将问题模型构建为马尔可夫决策过程，并设计基于分布式近端策略优化的节点选择算法，在每次训练迭代前选择合理的设备集合完成模型聚合。	Finally, the problem model was constructed as a Markov decision process and a node selection algorithm based on distributed proximal strategy optimization was designed to obtain a reasonable set of devices before each train- ing iteration to complete model aggregation.
3097	仿真实验表明，所提方法显著提高了联邦学习的准确率和训练速度，且具有良好的收敛性和稳健性。	Simulation results demonstrate that the proposed method significantly im- proves the accuracy and training speed of federated learning, and its convergence and robustness are also well.
3098	近年来，如何识别影响力最大的重要节点已成为网络科学最前沿的热点方向。	In recent years, how to select the most influential key node for identification has become the most cutting-edge hot direction in network science.
3099	将复杂网络节点影响力最大化问题表述为一个优化问题，其成本函数表示为节点影响力及其间的距离，使用 Shannon 熵对节点影响力进行度量，并利用一种改进灰狼优化算法来解决此问题。	Formulating the problem of maximizing the influence of complex network nodes as an optimization problem whose cost function was expressed as the influence of nodes and the distance between them, meas- ures user influence using Shannon entropy, and solved this problem using an improved gray wolf optimization algorithm.
3100	最后，使用真实复杂网络数据集进行数值计算。	Finally, numerical examples were performed with real complex network datasets.
3101	结果表明，与现有算法相比，所提算法精度更高，且计算效率较高。	The experimental results show that the proposed algorithm is more accurate and computationally efficient than the existing method.
3102	针对传统流形学习在数据降维时不考虑原数据类别和聚类程度低的缺陷，提出了一种有监督判别投影（SDP）的流形学习降维算法来改善网络安全数据降维效果。	In response to the problem that for dimensionality reduction, traditional manifold learning algorithm did not consider the raw data category information, and the degree of clustering was generally at a low level, a manifold learning dimensionality reduction algorithm with supervised discriminant projection (SDP) was proposed to improve the dimen- sionality reduction effects of network security data.