ID 原文 译文
58178 2 种算法仅与数据集的数据结构有关,不受其他外部参数影响. In order to solve the problem,two new hybrid algorithms based on fuzzy-rough instanceselection were proposed respectively,which are only related to intrinsic data structure of datasets and arenot affected by other external parameters.
58179 实验结果表明,基于模糊粗糙集实例选择的 2 种混合算法针对不同结构的数据集表现出了各自的特性,深化了对数据集的理解,提高了准确率. The experimental results show that the proposed hybrid algorithms exhibit their own characteristics for datasets with different structures,which deepens the understanding of data sets and improves the accuracy.
58180 为了克服传统语音端点检测算法在低信噪比环境下准确率低的问题,提出一种基于谱熵梅尔积( MFPH) 的语音端点检测算法. In order to solve the problem that the accuracy of traditional voice activity detection algorithmsis low in the low signal-to-noise ratio ( SNR) environment,a voice activity detection algorithm based onproduct of spectral entropy and Mel ( MFPH) was proposed.
58181 首先,提取带噪语音信号的梅尔频率倒谱系数中的第一维参数 MFCC0,将其与谱熵的乘积作为最终区分语音段和背景噪声段的融合特征参数; Firstly,the first dimensional parameterMFCC0 of Mel frequency spectrum coefficient of the speech signal with noisy was extracted,and the product of MFCC0 and spectral entropy was taken as fusion characteristic parameter of finally distinguishingspeech segment from background noise.
58182 然后,结合模糊 C 均值聚类算法和贝叶斯信息准则( BIC) 算法对MFPH 特征参数门限值进行自适应估计; Then,the threshold value of MFPH characteristic parameters wasestimated adaptively based on combination of fuzzy C-means clustering algorithm ( FCM) and Bayesianinformation criterion ( BIC) .
58183 最后,采用双门限法进行语音端点检测. Finally,the double-threshold method was adopted for the voice activitydetection.
58184 实验结果证明,与传统方法比较,该方法在 5 15 dB 低信噪比环境下的语音端点检测准确率有较大提高. Experiments show that the accuracy of the proposed method is greatly improved in the 5 ~15 dB low SNR environment compared with traditional methods.
58185 评估机会网络的关键节点可以发现对网络吞吐量影响最大的节点,为网络的优化和维护提供支撑. By evaluating critical nodes of opportunistic networks,it was found the nodes that have thegreatest influence on the throughput of network,which can support for network optimization and maintenance.
58186 为此,针对机会网络拓扑结构动态变化的特性构建了拓扑凝聚图,定义了二阶节点度、连接强度和关键域重要度 3 个评估指标,以指标的欧式距离表征节点的重要性. The topological condensation graph was constructed according to the characteristics of frequenttopology changes in opportunistic networks,and three evaluation metrics,such as second-order degree,connection strength,and key domain importance,were defined. The Euclidean distance of the metricswas employed to characterize the importance of the nodes.
58187 实验结果表明,与介数中心性方法相比,提出的模型具有有效性和优越性,并且模型在时间窗取 20 min 时具有较高的精度. Experiments show that the proposed model iseffective and superior compared with the betweenness method,and the model has higher accuracy whenthe time window is set for 20 minutes.