检索结果-内蒙古大学图书馆

International Joint Conference on Neural networks (IJCNN)

作者： Chuntao Ding Li Zhang Bangjun Wang School of Computer Science and Technology & Provincial Key laboratory for Computer Information Processing Soochow University Suzhou Jiangsu China

Discriminant neighborhood embedding (DNE) algorithm is one of supervised linear dimensionality reduction methods. Its nonlinear version kernel discriminant neighborhood embedding (KDNE) is expected to behave well on classification tasks. However, since KDNE constructs an adjacent graph in the original space, the adjacency graph could not represent the adjacent information in the kernel mapping space. By introducing hidden space, this paper proposes a novel nonlinear method for DNE, called hidden space discriminant neighborhood embedding (HDNE). This algorithm first maps the data in the original space into a high dimensional hidden space by a set of nonlinear hidden functions, and then builds an adjacent graph incorporating neighborhood information of the dataset in the hidden space. Finally, DNE is used to find a transformation matrix which would map the data in the hidden space to a low-dimensional subspace. The proposed method is applied to ORL face and MNIST handwritten digit databases. Experimental results show that the proposed method is efficiency for classification tasks.

关键词： Joints Conferences Neural networks

来源：评论

学校读者我要写书评

暂无评论

A modified approach of hot topics found on micro-blog

A modified approach of hot topics found on micro-blog

引用

5th International Symposium on IT in Medicine and Education, ITME 2013

作者： Ran, Lu Suzhi, Xue Yuanyuan, Ren Zhenfang, Zhu School of Information Science and Engineering Shandong Normal University Jinan 250014 China Shandong Provincial Key Laboratory for Novel Distributed Computer Software Technology Jinan 250014 China School of Information Science and Electric Engineering Shandong Jiaotong University Jinan 250357 China

ISBN: (纸本)9789400776173

Due to the simplicity, immediacy and convenience, micro-blog is gaining more and more attention from all kinds of people, especially the researchers. Recently, topic detection on micro-blog has attracted more interests due largely to the rapid development of micro-blog. However, retrieving information from micro-blog is challenging, as the texts of the micro-blog are short, ungrammatical, and unstructured, and they are full of noise. Therefore, the traditional hot topic detection method performed less. In order to solve this problem, this paper proposed a method of hot topics found based on speed growth. In this method, the pretreated micro-blogs were divided into different windows, and the time information was extracted in each window;then, for each word, it was expressed as feature trajectory of binary group sequence;then, calculated the growth speed of the word and the users relevant to the word in every adjacent two windows, selected the words whose growth speed is greater than a certain threshold as hot keywords;then, hot topics were found through the hot keywords clustering. The experiment was done based on SINA micro-blog dataset, the miss rate and false detection rate were done to prove the feasibility of the algorithm, results showed that the method improved the efficiency of the detection to a certain extent. © Springer science+Business Media Dordrecht 2014.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

AN ENSEMBLE OF DEEP NEURAL networks FOR OBJECT TRACKING

AN ENSEMBLE OF DEEP NEURAL NETWORKS FOR OBJECT TRACKING

引用

IEEE International Conference on Image Processing

作者： Xiangzeng Zhou Lei Xie Peng Zhang Yanning Zhang Shaanxi Provincial Key Laboratory of Speech & Image Information Processing (SAIIP) School of Computer Science Northwestern Polytechnical University

ISBN: (纸本)9781479957521

Object tracking in complex backgrounds with dramatic appearance variations is a challenging problem in computer vision. We tackle this problem by a novel approach that incorporates a deep learning architecture with an on-line AdaBoost framework. Inspired by its multi-level feature learning ability, a stacked denoising autoencoder (SDAE) is used to learn multi-level feature descriptors from a set of auxiliary images. Each layer of the SDAE, representing a different feature space, is subsequently transformed to a discriminative object/background deep neural network (DNN) classifier by adding a classification layer. By an on-line AdaBoost feature selection framework, the ensemble of the DNN classifiers is then updated on-line to robustly distinguish the target from the background. Experiments on an open tracking benchmark show promising results of the proposed tracker as compared with several state-of-the-art approaches.

关键词： Boosting Deep learning Visual tracking Deep neural network AdaBoost

来源：评论

学校读者我要写书评

暂无评论

ESYN: Efficient synchronization clustering algorithm based on dynamic synchronization model

引用

Tongxin Xuebao/Journal on Communications 2014年 35卷 86-93页

作者： Dong, Xue-Wen Yang, Chao Sheng, Li-Jie Ma, Jian-Feng Shaanxi Key Laboratory of Network and System Security Xidian University Xi'an710071 China School of Computer Science and Technology Xidian University Xi'an710071 China

Clustering is an important research field in data mining. Based on dynamical synchronization model, an efficient synchronization clustering algorithm ESYN is proposed. Firstly, based on local structure information of a non-vector network, a new concept vertex similarity is brought up to describe the link density between vertices. Secondly, the network is vectoried by OPTICS algorithm and turned into one-dimensional coordination sequence. Finally, global coupling analysis is applied to generalized Kuramoto synchronization model, synchronization radius is increased and the optimal clustering result is automatically selected. The experimental results on a large number of synthetic and real-world networks show that proposed algorithm achieves high accuracy. ©, 2014, Editorial Board of Journal on Communications. All right reserved.

关键词： Clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

A shared pixel ambient occlusion method in ray tracing

A shared pixel ambient occlusion method in ray tracing

引用

International Conference on Wavelet Active Media technology and Information Processing (ICWAMTIP)

作者： Yuxiao Guo Leiting Chen Tao Wu Xin Li Yue Cao School of Computer Science and Engineering University of Electronic Science and Technology of China Provincial Key Laboratory of Digital Media

ISBN: (纸本)9781479972098

Ambient occlusion is an illumination simulation approach used in ray tracing for decades. However, huge consume of multiple sample times, for a high re-solution, in ambient occlusion restricts its development. We propose an algorithm decreasing the sample times with adjacent pixels shared information in sampling process. By introducing this nova approach, we get a big improvement when comparing to previous methods.

关键词： Monte Carlo methods Ray tracing Graphics processing units Lighting Indexes Algorithm design and analysis Acceleration

来源：评论

学校读者我要写书评

暂无评论

A Framework for Workload Aware SaaS Platform Data Management

A Framework for Workload Aware SaaS Platform Data Management

引用

2014全国理论计算机科学学术年会

作者： Wei Guo Xinjun Wang Shandong Provincial Key Laboratory of Software Engineering School of Computer Science and TechnologyShandong UniversityChina

Cloud computing is a new emerging way by which hardware and software resources can be shared on-demand to computers and other devices via the *** computing platform has now become a new platform for enterprises and personal *** data placement issues in cloud computing platform have been extensively studied and researched in the past *** this paper,based on the workload information derived from system logs,we design an overall system architecture for the method of placing data replica and dynamic selection of data replica from a higher *** model the previous query workload as a hypergraph which contains a set of data items,then formulate and analyze the problem of replica placement by the graph theory *** use these concepts to develop a series of algorithms to decide which data items to replicate,and where to place the *** on this system architecture,we designed the workload-driven replica selection and placement *** to this algorithm,we can minimize the average query range of the involved *** experiments show that the fine-grained data placement and replication method can dramatically reduce the average query range resulting in significant reductions in the resource consumption in cloud computing platform.

关键词： Cloud computing Workload Data placement Data replica Replica selection

来源：评论

学校读者我要写书评

暂无评论

A model based on crossed cubes for VoD services

A model based on crossed cubes for VoD services

引用

International Conference on Systems and Informatics (ICSAI)

作者： Haifei Shen Jianxi Fan Baolei Cheng Cheng-Kuan Lin School of Computer Science and Technology Soochow University Suzhou China Provincial Key Laboratory for Computer Information Processing Technology Soochow University China

ISBN: (纸本)9781479954599

Content distribution schemes for Video on Demand (VoD) systems, based on the peer-to-peer (P2P) technology, have attracted more and more attention. Recently, people mainly focus on the latency performance, security, interaction, scalability, and so on. We propose a new network topology model - Similarity Crossed Cube(S-CQ for short), which combines with the feature of crossed cube and establishes multiple independent spanning trees for data distribution in the layer. This network model processes the properties such as good self-organization, delay, etc. The simulation results also show that it can effectively reduce the playback delay and maintain a low delay. The S-CQ network model can provide good quality of VoD service and effectively improve the user experience.

关键词： Routing Delays Bandwidth Streaming media Network topology Peer-to-peer computing Servers

来源：评论

学校读者我要写书评

暂无评论

Arranging the layout of alphanumeric buttons-the role of passwords 21

Arranging the layout of alphanumeric buttons-the role of pas...

引用

21st ACM Conference on computer and Communications Security, CCS 2014

作者： Li, Xiangxue Yu, Yu Li, Qiang Qian, Haifeng Zhou, Yuan Weng, Jian Department of Computer Science and Technology East China Normal University China State Key Laboratory of Integrated Services Networks Xidian University China School of Electronic Information and Electrical Engineering Shanghai Jiaotong University China Network Emergency Response Technical Team Coordination Center China Computer Science Department Jinan University China

ISBN: (纸本)9781450329576

A typical but trivial layout of alphanumeric buttons in the touchscreen setting is to arrange the 10 digits and 26 letters in a natural order. This arrangement does not take into account the frequencies of letters and digits when the users touch the buttons to key in their passwords or messages. We examine large scale datasets of over 141 million passwords collected from several leading websites for social networking, Internet forums, gaming, dating, and various other online service providers in China, and find that the distribution of letters in passwords is quite close to that in Chinese language. Based on the letter/digit frequencies, we further propose an alphanumeric button layout scheme with the following advantages: the buttons are clicked as uniformly as possible, so that the lifetime of the touchscreen can be prolonged and finger oil residues may scatter more evenly over the button area of the screen;and in the meantime, the movements of users' fingers are improved to enhance good user experience when inputting messages. The idea behind the layout is potentially applicable to diversified races. Copyright is held by the author/owner(s).

关键词： Fasteners

来源：评论

学校读者我要写书评

暂无评论

Strategy of domain and cross-domain access control based on trust in cloud computing environment

Strategy of domain and cross-domain access control based on ...

引用

3rd International Conference on computer Engineering and Network, CENet 2013

作者： Li, Bo Tian, Ming Zhang, Yongsheng Lv, Shenjuan Academic Affairs Office Shandong Polytechnic Jinan 250104 China School of Information Science and Engineering Shandong Normal University Jinan 250014 China Shandong Provincial Key Laboratory for Novel Distributed Computer Software Technology Jinan 250014 China

ISBN: (纸本)9783319017655

Under the current cloud computing environment, a reasonable and practicable access control strategy is needed, which is a guarantee to protect cloud computing suppliers to provide services and many cloud users access to services. In this paper, based on analysis of many cloud computing safety features, trust management is introduced into the cloud computing service access control, within the domain of a trust-based access control strategy, in domain, presents a trust-based access control policy. Credible value will be given through the comprehensive treatment of the entity, and then AAC (authentication and authorization center) authorizes the appropriate access rights to achieve the control of the monomer in the domain. Combined with the characteristics of the existing cloud computing environment, in multiple management domains, this paper proposes a role mapping, with the role mapping relationship between the domain, which can make the inter-domain access to resources and security shared access between different domains, in order to avoid the problem of permission penetration and privilege escalation, this paper presents the mirror role based on role mapping, ultimately solves the problem. © 2014 Springer International Publishing Switzerland.

关键词： Cloud computing

来源：评论

学校读者我要写书评

暂无评论

The application of trusted computing technology in the cloud security

The application of trusted computing technology in the cloud...

引用

3rd International Conference on computer Engineering and Network, CENet 2013

作者： Li, Bo Lv, Shenjuan Zhang, Yongsheng Tian, Ming Academic Affairs Office Shandong Polytechnic Jinan 250104 China School of Information Science and Engineering Shandong Normal University Jinan 250014 China Shandong Provincial Key Laboratory for Novel Distributed Computer Software Technology Jinan 250014 China

ISBN: (纸本)9783319017655

For the lack of safety and reliability of the information in the cloud computing environment, in order to create a more flexible and adaptable security mechanism, the combination of cloud computing and credible concept is a major research direction in today's security. Based on the view mentioned above, this paper strengthens the research of trust computing technology to solve the security issues in cloud and cloud-based trust transfer, on the basis of the practical work of the experts and scholars on the trust transfer technology, and expands the theoretical model of the trust chain. This paper uses the stochastic process algebra and Petri nets as a modeling tool to build two trust chain models, demonstrates the credibility of certain behavioral characteristics of the chain, analyzes several constraints of credible chain, and provides a valuable reference for engineering practice of the credible chain. © 2014 Springer International Publishing Switzerland.

关键词： Chains

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：