检索结果-内蒙古大学图书馆

2022 Conference on Research in Adaptive and Convergent Systems, RACS 2022

作者： Zhou, Wenxuan Qu, Zhihao Zhao, Yanchao Tang, Bin Ye, Baoliu Key Laboratory of Water Big Data Technology of Ministry of Water Resources and School of Computer and Information Hohai University Jiangsu Nanjing China State Key Laboratory for Novel Software Technology Nanjing University Jiangsu Nanjing China College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Jiangsu Nanjing China

ISBN: (纸本)9781450393980

Split learning is an emerging distributed machine learning paradigm that exhibits great advantages in privacy protection and training efficiency. Most previous studies pay attention to client-server based split learning on Convolutional Neural networks (CNNs), and little effort has been spent on investigating the Recurrent Neural Networks (RNNs) oriented split learning. To fill in this gap, we propose a multi-hop split learning algorithm for RNN in mobile edge computing, namely M-SLRNN. The core idea is to efficiently orchestrate resource-constrained mobile devices to collaboratively train a large-size model. We formulate the optimization problem with the objective of minimizing the total training time and theoretically analyze its hardness. Then, we propose an efficient algorithm that conditionally has an approximation ratio of 2. Extensive experiments on both image and text classification tasks demonstrate the effectiveness and the efficiency of the proposed algorithm. © 2022 ACM.

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

Speech Intelligibility Enhancement Algorithm Based on Multi-Resolution Power-Normalized Cepstral Coefficients(MRPNCC)for Digital Hearing Aids

引用

computer Modeling in Engineering & Sciences 2021年第2期126卷 693-710页

作者： Xia Wang Xing Deng Hongming Shen Guodong Zhang Shibing Zhang School of Information Science and Technology Nantong UniversityNantong226019China School of Automation Key Laboratory of Measurement and Control for CSEMinistry of EducationSoutheast UniversityNanjing210096China School of Computer Science Jiangsu University of Science and TechnologyZhenjiang212003China

Speech intelligibility enhancement in noisy environments is still one of the major challenges for hearing impaired in everyday ***,Machine-learning based approaches to speech enhancement have shown great promise for improving speech *** key issues of these approaches are acoustic features extracted from noisy signals and classifiers used for supervised *** this paper,features are ***-resolution power-normalized cepstral coefficients(MRPNCC)are proposed as a new feature to enhance the speech intelligibility for hearing *** new feature is constructed by combining four cepstrum at different time–frequency(T–F)resolutions in order to capture both the local and contextual *** vectors and binary masking labels calculated by signals passed through gammatone filterbank are used to train support vector machine(SVM)classifier,which aim to identify the binary masking values of the T–F units in the enhancement *** enhanced speech is synthesized by using the estimated masking values and wiener filtered T–F *** experimental results demonstrate that the proposed feature is superior to other comparing features in terms of HIT-FA,STOI,HASPI and PESQ,and that the proposed algorithm not only improves speech intelligibility but also improves speech quality *** tests validate the effectiveness of the proposed algorithm for hearing impaired.

关键词： Speech intelligibility enhancement multi-resolution power-normalized cepstral coefficients binary masking value hearing impaired

来源：评论

学校读者我要写书评

暂无评论

Unsupervised ensemble learning based on graph embedding for image clustering 25th

Unsupervised ensemble learning based on graph embedding for ...

引用

25th International Conference on Neural information processing, ICONIP 2018

作者： Luo, Xiaohui Zhang, Li Li, Fanzhang Hu, Chengxiang School of Computer Science and Technology and Joint International Research Laboratory of Machine Learning and Neuromorphic Computing Provincial Key Laboratory for Computer Information Processing Technology Soochow University SuzhouJiangsu215006 China

ISBN: (纸本)9783030041816

Manifold learning has attracted more and more attention in machine learning for past decades. Unsupervised Large Graph Embedding (ULGE), which performs well on the large-scale data, has been proposed for manifold learning. To improve the clustering performance, a novel Unsupervised Ensemble Learning based on Graph Embedding (UEL-GE) is explored, which takes ULGE to get low-dimensional embeddings of the given data and uses the K-means method to obtain the clustering results. Furthermore, the multiple clusterings are corrected by using the bestMap method. Finally, the corrected clusterings are combined to generate the final clustering. Extensive experiments on several data sets are conducted to show the efficiency and effectiveness of the proposed ensemble learning method. © Springer Nature Switzerland AG 2018.

关键词： K-means clustering

来源：评论

学校读者我要写书评

暂无评论

A Forward-Secure Multi-Receiver Signcryption Scheme

引用

Journal of Harbin Institute of technology(New Series) 2016年第5期23卷 53-61页

作者： Zhimin Yu Zhengjun Jing Hua Yang Chunsheng Gu Key Laboratory of Cloud Computing and Intelligent Information Processing of Changzhou City Jiangsu University of Technology Changzhou Jiangsu 213001 China College of Computer Nanjing University of Posts and Telecommunications Nanjing 210003 China State Key Laboratory of Information Security Institute of Information Engineering Chinese Academy of Sciences Beijing 100093 China

In order to protect historical ciphertext when the private key leaked in the broadcasting system,the forward-secure multi-receiver signcryption scheme is designed based on the generic graded multilinear mapping encoding structure,which effectively prevents illegal access from intruder to the ciphertext in the past time period when the private key in current time period is *** the generalization of the existing multilinear mapping encoding system,it proposes the generic graded multilinear mapping encoding structure and the generic graded decision Diffie-Hellman *** of the generic graded multilinear mapping encoding system adopted,almost all candidate multilinear mapping encoding systems can automatically adapt to our *** the assumption of generic graded decision Diffie-Hellman problem,it has proved that the scheme has the information confidentiality and unforgeability in the current time *** putting forward the security model of forward-secure multi-receiver signcryption scheme,and under the assumption of generic graded decision Diffie-Hellman problem,it has proved that the scheme has the message forward-confidentiality and *** with other forward-secure public key encryption schemes,the relationship between our scheme and time periods is sub-linear,so it is less complex.

关键词： multilinear mapping forward-secure multi-receiver signcryption confidentiality unforgeability

来源：评论

学校读者我要写书评

暂无评论

Virtual brain surgery simulation system based on haptic interaction

引用

High technology Letters 2015年第2期21卷 185-191页

作者：张小瑞 Zhu Jiandong Sun Wei Norman I.Badler Song Aiguo Niu Jianwei Liu Jia School of Computer and Software Nanjing University of Information Science & Technology Jiangsu Engineering Center of Network Monitoring Nanjing University of Information Science and Technology School of Information and Control Nanjing University of Information Science & Technology School of Engineering and Applied Science University of Pennsylvania School of Instrument Science and Engineering Southeast University State key Laboratory of Virtual Reality Technology and Systems Beihang University

To improve the accuracy and interactivity of soft tissue delormatlon simulation, a new plate spring model based on physics is proposed. The model is parameterized and thus can be adapted to simulate different organs. Different soft tissues are modeled by changing the width, number of pieces, thickness, and length of a single plate spring. In this paper, the structural design, calcula- tion of soft tissue deformation and real-time feedback operations of our system are also introduced. To evaluate the feasibility of the system and validate the model, an experimental system of haptic in- teraction, in which users can use virtual hands to pull virtual brain tissues, is built using PHANTOM OMNI devices. Experimental results show that the proposed system is stable, accurate and promising for modeling instantaneous soft tissue deformation.

关键词： haptic feedback human-computer interaction surgery simulation soft tissue deformation

来源：评论

学校读者我要写书评

暂无评论

Simulation on nanoimprint of grating structures

Simulation on nanoimprint of grating structures

引用

2010 International Conference on computer Application and System Modeling, ICCASM 2010

作者： Sun, Hongwen Liu, Guogao Lin, Shanming College of Computer and Information Engineering Hohai University Changzhou China Jiangsu Key Laboratory of Power Transmission and Distribution Equipment Technology China

ISBN: (纸本)9781424472369

Nanoimprint lithography (NIL) is an important nanolithography process with low cost, high throughput and high resolution. It creates patterns by mechanical deformation of imprint resist. Fabricating gratings by NIL can manufacture lots of replica in a fast way. Simulation on the NIL process has a vital role on choosing optimized parameters. One finite element analysis software DEFORM was used to analyze different imprint factors. Through simulation optimization, it was found that imprint temperature, pressure and time strongly affected the replication fidelity. The simulation result will guide the NIL experiment. © 2010 IEEE.

关键词： Nanoimprint lithography

来源：评论

学校读者我要写书评

暂无评论

EFFECTS OF DIFFERENT FUNCTIONALIZATION SCHEMES ON THE INTERFACIAL STRENGTH OF CARBON NANOTUBE POLYETHYLENE COMPOSITE

引用

Acta Mechanica Solida Sinica 2015年第3期28卷 277-284页

作者： Tan Xiao Juqing Liu Huifang Xiong Architecture and Civil Engineering Institute Guangdong University of Petrochemical Technology Key Laboratory of Flexible Electronics (KLOFE) & Institute of Advanced Materials (IAM) Jiangsu National Synergistic Innovation Center for Advanced Materials (SICAM)Nanjing Tech University (NanjingTech) School of Computer and Electronic Information Guangdong University of Petrochemical Technology

The mechanical performance of carbon nanotube （CNT） reinforced polymer com- posites is primarily controlled by the dispersive capacity and interracial shear strength of CNTs in polymer matrices. CNT functionalizations will improve dispersion and strengthen interracial bonding of CNTs in matrices. To understand the effects of different functionalization schemes on the interracial strength of CNT-polymer composites, pullout of the covalent, noncovalent, and mixed functionalized single-walled carbon nanotube （SWCNT） from polyethylene （PE） matrix was simulated by using molecular dynamics, respectively. The results show that the SWCNT-PE interracial shear strength is significantly improved by SWCNT functionalizations, particularly by mixed functionalization.

关键词： interracial shear strength carbon nanotubes nano composites molecular dynamics

来源：评论

学校读者我要写书评

暂无评论

Specification and counterexample generation for cyber-physical systems

引用

Soft Computing 2024年第17-18期28卷 9137-9155页

作者： Li, Zhen Cao, Zining Wang, Fujun Xing, Chao College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Jiangsu Nanjing211106 China Science and Technology on Electro-optic Control Laboratory Henan Luoyang471000 China Collaborative Innovation Center of Novel Software Technology and Industrializaton Nanjing University of Aeronautics and Astronautics Jiangsu Nanjing210023 China Key Laboratory of Safety-Critical Software Ministry of Industry and Information Technology Jiangsu Nanjing211106 China

Cyber-Physical Systems (CPS) are complex systems that integrate information control devices with physical resources, which can be automatically and formalized verified by model checking according to the expected requirements in the formal specification. The counterexamples in model checking are witnesses to the violation of the specification properties of the system and can provide important diagnostic information for debugging, controlling, and synthesizing CPS. Designing a rational specification language for CPS and generating effective counterexamples allows security vulnerabilities to be detected and addressed early in the system development. However, CPS involve frequent interactions between cyber and physical systems and often operate in unreliable environments, which poses new challenges for comprehensive modeling and designing specification languages for CPSs with discrete, continuous, time, probabilistic, and concurrent behaviors. Moreover, finding the smallest counterexample of CPS with probabilistic behavior in the shortest possible time has been identified as a Non-Deterministic Polynomial-complete (NP-complete) problem. Although a number of heuristics have been devised to address this challenge, the accuracy and efficiency of the solved counterexamples need to be improved due to the difficulty in determining the heuristic functions. We first provide a comprehensive model for CPS by introducing the Hybrid Probabilistic Time Labeled Transition System (HPTLTS). Subsequently, we design a specification language HPTLTS Temporal Logic (HPTLTS-TL) that can describe the properties of CPS. In addition, we propose an optimization algorithm CACO-A, which combines the Ant Colony Optimization (ACO) algorithm and the A-algorithm to efficiently generate the counterexample of CPS, which is represented as the diagnostic subgraph. Finally, we discuss a typical CPS example to demonstrate the feasibility of our approach. © The Author(s), under exclusive licence to Springer-

关键词： Embedded systems

来源：评论

学校读者我要写书评

暂无评论

Inducing bilingual lexica from non-parallel data with earth mover's distance regularization 26

Inducing bilingual lexica from non-parallel data with earth ...

引用

26th International Conference on Computational Linguistics, COLING 2016

作者： Zhang, Meng Liu, Yang Luan, Huanbo Liu, Yiqun Sun, Maosong State Key Laboratory of Intelligent Technology and Systems Tsinghua National Laboratory for Information Science and Technology Department of Computer Science and Technology Tsinghua University Beijing China Jiangsu Collaborative Innovation Center for Language Competence Jiangsu China

ISBN: (纸本)9784879747020

Being able to induce word translations from non-parallel data is often a prerequisite for cross-lingual processing in resource-scarce languages and domains. Previous endeavors typically simplify this task by imposing the one-to-one translation assumption, which is too strong to hold for natural languages. We remove this constraint by introducing the Earth Mover's Distance into the training of bilingual word embeddings. In this way, we take advantage of its capability to handle multiple alternative word translations in a natural form of regularization. Our approach shows significant and consistent improvements across four language pairs. We also demonstrate that our approach is particularly preferable in resource-scarce settings as it only requires a minimal seed lexicon. © 1963-2018 ACL.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Mining approximate keys based on reasoning from XML data

Mining approximate keys based on reasoning from XML data

引用

17th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2013

作者： Yijun, Liu Feiyue, Ye Sheng, He School of Computer Engineering Jiangsu University of Technology Changzhou Jiangsu 213001 China Key Laboratory of Cloud Computing and Intelligent Information Processing of Changzhou City Changzhou Jiangsu 213001 China

ISBN: (纸本)9783642403187

keys are very important for data management. Due to the hierarchical and flexible structure of XML, mining keys from XML data is a more complex and difficult task than from relational databases. In this paper, we study mining approximate keys from XML data, and define the support and confidence of a key expression based on the number of null values on key paths. In the mining process, inference rules are used to derive new keys. Through the two-phase reasoning, a target set of approximate keys and its reduced set are obtained. Our research conducted experiments over ten benchmark XML datasets from XMark and four files in the UW XML Repository. The results show that the approach is feasible and efficient, with which effective keys in various XML data can be discovered. © Springer-Verlag 2013.

关键词： Data mining

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：