ID	原文	译文
4793	该算法借鉴极端梯度提升（XGBoost）算法中构建树的思想过程，通过从 3 个重要性度量的角度来衡量特征的重要性，避免单一重要性度量的局限性；	The thought process of building trees inXGBoost was used for reference, and the importance of features from three importance metrics was measured to avoidthe limitation of single importance metric.
4794	然后通过改进的序列浮动前向搜索策略（ISFFS）搜索特征子集，使最终得到的特征子集有较高的质量。	Then the improved sequential floating forward selection (ISFFS) was appliedto search the feature subset so that it had high quality.
4795	在 8 个 UCI 数据集的对比实验中表明，所提算法具有很好的性能。	Compared with the experimental results of eight datasets in UCI,the proposed algorithm has good performance.
4796	识别重要节点是复杂网络研究的基础性问题。	Identifying vital nodes is a basic problem in complex network research.
4797	现有理论框架主要以“点−边”这种低阶结构为基本单元，往往忽略了多个节点之间可能存在的交互性、传递性等重要因素。	The existing theoretical framework,mainly considered from the lower-order structure of node-based and edge-based relations often ignores important factorssuch as interactivity and transitivity between multiple nodes.
4798	为了更加精确地识别重要节点，对网络中以模体为基本单元的高阶结构进行了研究，首先，提出了节点高阶度的概念，进一步引入证据理论融合了节点的高阶结构和低阶结构信息，设计了一种融合节点高阶信息的半局部重要节点识别方法。	To identify vital nodes more accurately, the motif, thehigh-er-order structure of the network, was studied as the basic unit. Firstly, a notion of higher-order degree of nodes in acom-plex network was proposed. Then, the higher-order structure and lower-order structure of nodes were fused intoevidence theory.
4799	在 3 个真实社交网络上的实验结果表明，相较于只关注低阶结构的已有方法，所提出的算法能够更加精确地识别网络中的重要节点。	A semi-local identifying vital nodes algorithm fusing higher-order information of nodes was designed.The results of experiments on three real social networks show that the proposed algorithm can identify vital nodes moreaccurately in the network than the existing methods which only focus on the low-order structure.
4800	针对主动声呐中回波信号特征提取困难的问题，提出了一种利用降噪自编码器与卷积降噪自编码器相结合的自编码器算法。	Aiming at the difficulty of feature extraction of echo signal in active sonar, a self-encoder algorithm based on thecombination of denoising self-encoder and convolution denoising self-encoder was proposed.
4801	首先利用降噪自编码器在信号整体上的降噪优势，对含噪信号进行预处理；然后结合卷积降噪自编码器对信号局部特征的优化，对信号进行局部降噪，从而实现信号增强。	Firstly, the preprocessing ofnoisy signal was carried out by using the advantage of denoising self-encoder in signal as a whole, and then the local featureof signal was optimized by combining convolutional denoising self-encoder to denoise the signal locally, so as to enhance thesignal.
4802	所提算法直接以接收信号的时域波形作为特征输入，保留了信号的幅度与相位特征。	The time domain waveform of the received signal is used as the feature input by the algorithm, and retains the signal'samplitude and phase characteristics.