检索结果-内蒙古大学图书馆

IEEE International Conference on Cognitive Informatics

作者： Zhao, Peng Geng, Huan-Tong Cai, Qing-Sheng Department of Computer Science and Technology University of Science and Technology of China Hefei 230027 China Key Lab. of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei 230039 China

ISBN: (纸本)9781424404759

After analyzing the disadvantages of traditional text clustering method based on keywords set, a novel approach for clustering of Chinese text based on concept hierarchy is presented. It introduces a Chinese topic classify dictionary as background knowledge to clustering of Chinese text. It adopts a hierarchical coding system which reflects concept relevance among different words and uses Vector Space Model based on concept hierarchy to represent Chinese text. The experimental results show this approach is more effective than traditional text clustering method based on keywords set. © 2006 IEEE.

关键词： Cluster analysis

来源：评论

学校读者我要写书评

暂无评论

SES: Bridging the Gap between Explainability and Prediction of Graph Neural Networks 40

SES: Bridging the Gap between Explainability and Prediction ...

引用

40th IEEE International Conference on Data Engineering, ICDE 2024

作者： Huang, Zhenhua Li, Kunhao Wang, Shaojie Jia, Zhaohong Zhu, Wentao Mehrotra, Sharad Anhui University Hefei China Key Lab of Intelligent Computing and Signal Processing of Ministry of Education Hefei China Amazon Research Seattle United States University of California Irvine Irvine United States

ISBN: (纸本)9798350317152

Despite the Graph Neural Networks' (GNNs) pro-ficiency in analyzing graph data, achieving high-accuracy and interpretable predictions remains challenging. Existing GNN interpreters typically provide post-hoc explanations disjointed from GNNs' predictions, resulting in misrepresentations. Self-explainable GNNs offer built-in explanations during the training process. However, they cannot exploit the explanatory outcomes to augment prediction performance, and they fail to provide high-quality explanations of node features and require additional processes to generate explainable subgraphs, which is costly. To address the aforementioned limitations, we propose a self-explained and self-supervised graph neural network (SES) to bridge the gap between explainability and prediction. SES comprises two processes: explainable training and enhanced predictive learning. During explainable training, SES employs a global mask generator co-trained with a graph encoder and directly produces crucial structure and feature masks, reducing time consumption and providing node feature and subgraph explanations. In the enhanced predictive learning phase, mask-based positive-negative pairs are constructed utilizing the ex-planations to compute a triplet loss and enhance the node representations by contrastive learning. Extensive experiments demonstrate the superiority of SES on multiple datasets and tasks. SES outperforms baselines on real-world node classification datasets by notable margins of up to 2.59% and achieves state-of-the-art (SOTA) performance in explanation tasks on synthetic datasets with improvements of up to 3.0%. Moreover, SES delivers more coherent explanations on real-world datasets, has a fourfold increase in Fidelity+ score for explanation quality, and demonstrates faster training and expla-nation generating times. To our knowledge, SES is a pioneering GNN to achieve SOTA performance on both explanation and prediction tasks. © 2024 IEEE.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Subspace selection using semi-supervised harmonic mean of Kullback-Leibler divergences

Subspace selection using semi-supervised harmonic mean of Ku...

引用

201. 7th International Conference on Fuzzy Systems and Knowledge Discovery, FSKD 201.

作者： Chen, Si-Bao Wang, Hai-Xian Luo, Bin Key Lab of Intelligent Computing and Signal Processing of Ministry of Education School of Computer Science and Technology Anhui University Hefei Anhui 230039 China Key Lab of Child Development and Learning Science of Ministry of Education Research Center for Learning Science Southeast University Nanjing Jiangsu 210096 China

ISBN: (纸本)9781424459346

In many areas of pattern recognition and machine learning, subspace selection is an essential step. Fisher's linear discriminant analysis (LDA) is one of the most well-known linear subspace selection methods. However, LDA suffers from the class separation problem. The projection to a subspace tends to merge close class pairs. A recent result, named maximizing the geometric mean of Kullback-Leibler (KL) divergences of class pairs (MGMD), can significantly reduce the class separation problem. Furthermore, maximizing the harmonic mean of Kullback-Leibler (KL) divergences of class pairs (MHMD) emphasizes smaller divergences more than MGMD, and deals with the class separation problem more effectively. However, in many applications, labeled data are very limited while unlabeled data can be easily obtained. The estimation of divergences of class pairs is unstable using inadequate labeled data. To take advantage of unlabeled data for subspace selection, semi-supervised MHMD (SSMHMD) is proposed using graph Laplacian as normalization. Quasi-Newton method is adopted to solve the optimization problem. Experiments on synthetic data and real image data show the validity of SSMHMD. ©201. IEEE.

关键词： Discriminant analysis

来源：评论

学校读者我要写书评

暂无评论

A full-rate wireless cooperative relay transmission scheme with low detection complexity

引用

Journal of Computational Information Systems 2012年第13期8卷 5279-5285页

作者： Zhang, Hongwei Liu, Yanru Zhang, Ye Li, Xiaohui Key Lab. of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei 230039 China

To achieve low detection complexity and full-rate wireless cooperative transmissions, pre-coding and cyclic delay diversity (CDD) techniques are fully considered and properly combined for a two-path successive relay network with multiple relay nodes in this paper. Firstly, the symbols to be transmitted at source node are encoded with linear complex field coding methods, and then these coded symbols are divided into two data streams. Both streams are transmitted to two groups of relay nodes respectively during two consecutive time slots. After having been amplified and forwarded by relay nodes using CDD ways, all transmit signals are received by the destination node finally. At receiver, only two symbols are needed to be jointly detected in each transmission process, the complexity is low. Simulation results show that the proposed scheme can achieve full-rate cooperative relay transmissions and obtain high diversity gains of the distributed wireless systems. © 201. by Binary Information Press.

关键词： Orthogonal frequency division multiplexing

来源：评论

学校读者我要写书评

暂无评论

A full-rate and low-detection complexity cooperative transmission scheme with cyclic delay diversity

引用

Journal of Computational Information Systems 2013年第11期9卷 4593-4600页

作者： Xu, Lei Wang, Li Zhang, Hongwei Li, Xiaohui Key Lab. of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei 230039 China

The use of orthogonal channels for the cooperative transmission results in a loss of rate or spectral efficiency, and the exiting full-rate cooperative transmission schemes based on space-time code design have the defect of high detection complexity. Therefore, by adopting the non-orthogonal amplify and forward (NAF) mode, combining with the advantages of linear constellation precoding (LCP) technique and cyclic delay diversity (CDD) technique in this paper, a new full-rate wireless cooperative transmission scheme designed based on space-time-frequency codes is proposed. Compared with the existing full-rate cooperative transmission scheme, this new scheme is suitable for multi-relay scenario, and has the advantages of simple structure, low signal detection complexity which does not increase with the relay number. Compared with the equal power allocation scheme, the optimal power allocation scheme can make the system bit error rate (BER) performance improved. © 201. by Binary Information Press.

关键词： Spectrum efficiency

来源：评论

学校读者我要写书评

暂无评论

Speech Endpoint Detection in Noisy Environments Using EMD and Teager Energy Operator

引用

Journal of Electronic Science and Technology 2010年第2期8卷 183-186页

作者： De-Xiang Zhang Xiao-Pei Wu Zhao Lv Key Lab. of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei 230039 China

Accurate endpoint detection is a necessary capability for speech recognition. A new energy measure method based on the empirical mode decomposition （EMD） algorithm and Teager energy operator （TEO） is proposed to locate endpoint intervals of a speech signal embedded in noise. With the EMD, the noise signals can be decomposed into different numbers of sub-signals called intrinsic mode functions （IMFs）, which is a zero-mean AM-FM component. Then TEO can be used to extract the desired feature of the modulation energy for IMF components. In order to show the effectiveness of the proposed method, examples are presented to show that the new measure is more effective than traditional measures. The present experimental results show that the measure can be used to improve the performance of endpoint detection algorithms and the accuracy of this algorithm is quite satisfactory and acceptable.

关键词： Index Terms----Empirical mode decomposition endpoint detection noisy speech Teager energy operator.

来源：评论

学校读者我要写书评

暂无评论

Analysis of electromagnetic characteristics of antenna arrays based on the characteristic basis function method

Analysis of electromagnetic characteristics of antenna array...

引用

International Symposium on Antennas Propagation and EM Theory

作者： Meng Wu Yufa Sun Guohua Wang Key Lab. of Intelligent Computing & Signal Processing Ministry of Education Anhui University

ISBN: (纸本)9781467317993

The characteristic basis function method (CBFM) is used to analyze the electromagnetic characteristics of antenna arrays in this paper. The original antenna arrays are divided into a number of subdomains and high-level basis functions are constructed for each subdomain through their couplings, the resulting matrix can be largely reduced. The characteristic basis function method not only has high accuracy and computing time is greatly reduced. The numerical results are given to illustrate that the method is accurate and efficient.

关键词： Characteristic basis function method Method of Moments Antenna arrays

来源：评论

学校读者我要写书评

暂无评论

A full-rate cooperative relay transmission scheme with randomized cyclic delay diversity

引用

Journal of Computational Information Systems 2014年第1期10卷 275-284页

作者： Xu, Lei Bao, Menglin Jiang, Ran Zhang, Hongwei Li, Xiaohui Key Lab. of Intelligent Computing and Signal Processing of Ministry of Education Anhui University Hefei 230039 China

Randomized cyclic delay diversity (RCDD) is an effective means to capture both the space diversity and the frequency diversity with low complexity detection over frequency-selective fading channels. Since many existing full-rate cooperative relay transmission schemes have the disadvantages of low spectrum efficiency and high complexity detection, a practical and low complexity detection full-rate distributed cooperative relay transmission scheme is proposed in this paper by taking advantages of the randomized cyclic delay diversity method and the linear constellation precoding (LCP) technology. The bit error rate (BER) performance of the proposed scheme is superior to that of the system with fixed cyclic delay diversity (FCDD) when the number of subcarriers is sufficiently large. Moreover, the system diversity gain improves with the multipath number of the frequency-selective fading channel and the relay number increases. Copyright © 201. Binary Information Press.

关键词： Cooperative communication

来源：评论

学校读者我要写书评

暂无评论

Fast calculation of bistatic RCS for conducting objects using the adaptive cross approximation algorithm

Fast calculation of bistatic RCS for conducting objects usin...

引用

International Symposium on Antennas Propagation and EM Theory

作者： Lei Chen Yufa Sun Shuaishuai Yang Key Lab. of Intelligent Computing & Signal Processing Ministry of Education Anhui University

ISBN: (纸本)9781467317993

Adaptive cross approximation (ACA) algorithm as a kind of common matrix compression method is often used to analyze the electromagnetic radiation and scattering problems. In the region of far field, the impedance matrix is compressed by ACA method, and the extracted impedance matrix is accelerated filling by equivalent dipole-moment method (EDM). In the area of near field, the method of moments or the equivalent dipole-moment method is applied for filling the impedance matrix. The iterative near field preconditioning technique is used to solve matrix equation, so that fast calculation of radar cross section (RCS) can be achieved. Numerical results show that the computational efficiency is improved significantly via applying the presented method in this paper without sacrificing much accuracy.

关键词： Adaptive cross approximation algorithm Equivalent dipole-moment method Near field preconditioning Radar cross section

来源：评论

学校读者我要写书评

暂无评论

SPEECH ENDPOINT DETECTION IN NOISY ENVIRONMENTS USING EMD AND TEAGER ENERGY OPERATOR

引用

Journal of Electronic Science and Technology 2010年第2期000卷 183-186页

作者： DE-XIANG ZHANG XIAO-PEI WU ZHAO Lü Key Lab.of Intelligent Computing and Signal Processing of Ministry of Education.Anhui University Hefei 230039China

Accurate endpoint detection is a necessary capability for speech recognition. A new energy measure method based on the empirical mode decomposition (EMD) algorithm and Teager energy operator (TEO) is proposed to locate endpoint intervals of a speech signal embedded in noise. With the EMD, the noise signals can be decomposed into different numbers of sub-signals called intrinsic mode functions (IMFs), which is a zero-mean AM-FM component. Then TEO can be used to extract the desired feature of the modulation energy for IMF components. In order to show the effectiveness of the proposed method, examples are presented to show that the new measure is more effective than traditional measures. The present experimental results show that the measure can be used to improve the performance of endpoint detection algorithms and the accuracy of this algorithm is quite satisfactory and acceptable.

关键词： Empirical mode decomposition endpoint detection noisy speech Teager energy operator

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：