检索结果-内蒙古大学图书馆

A survey on dependability improvement techniques for pervasive computing systems

science China(Information sciences) 2015年第5期58卷 19-32页

作者： YANG WenHua LIU YePang XU Chang CHEUNG S.C. State Key Laboratory for Novel Software Technology Nanjing University Department of Computer Science and Technology Nanjing University Department of Computer Science and Engineering The Hong Kong University of Science and Technology

The goal of this survey is to summarize the state-of-the-art research results and identify research challenges of developing and deploying dependable pervasive computing systems. We discuss the factors that affect the system dependability and the studies conducted to improve it with respect to these factors. These studies were categorized according to their similarities and differences in hope of shedding some insight into future research. There are three categories: context management, fault detection, and uncertainty *** three categories of work address the three most difficult problems of pervasive computing systems. First,pervasive computing systems’ perceived environments, which are also called their contexts, can vary intensively,and thus have a great impact on the systems’ dependability. Second, it is challenging to guarantee the correctness of the systems’ internal computations integrated with interactions with external environments for *** detection is then an important issue for improving dependability for these systems. Last but not least importantly, pervasive computing systems interact with their environments frequently. These interactions can be affected by many uncertainties, which can jeopardize the systems’ dependability. After a discussion of these pieces of work, we present an outlook for its future research directions.

关键词： pervasive computing system dependability context fault detection uncertainty

来源：评论

学校读者我要写书评

暂无评论

Learning Compact Binary Codes from Higher-Order Tensors via Free-Form Reshaping and Binarized Multilinear PCA

Learning Compact Binary Codes from Higher-Order Tensors via ...

引用

International Joint Conference on Neural Networks

作者： Haiping Lu Jianxin Wu Yu Zhang Department of Computer Science Hong Kong Baptist University National Key Lab for Novel Software Technology Nanjing University Bioinformatics Institute A*STAR

ISBN: (纸本)9781509006212

For big, high-dimensional dense features, it is important to learn compact binary codes or compress them for greater memory efficiency. This paper proposes a Binarized Multilinear PCA (BMP) method for this problem with Free-Form Reshaping (FFR) of such features to higher-order tensors, lifting the structure-modelling restriction in traditional tensor models. The reshaped tensors are transformed to a subspace using multilinear PCA. Then, we unsupervisedly select features and supervisedly binarize them with a minimum-classification-error scheme to get compact binary codes. We evaluate BMP on two scene recognition datasets against state-of-the-art algorithms. The FFR works well in experiments. With the same number of compression parameters (model size), BMP has much higher classification accuracy. To achieve the same accuracy or compression ratio, BMP has an order of magnitude smaller number of compression parameters. Thus, BMP has great potential in memory-sensitive applications such as mobile computing and big data analytics.

关键词： Tensile stress Binary codes Principal component analysis Image coding Three-dimensional displays Electronic mail Big data

来源：评论

学校读者我要写书评

暂无评论

Representation learning of knowledge graphs with hierarchical types 25

Representation learning of knowledge graphs with hierarchica...

引用

25th International Joint Conference on Artificial Intelligence, IJCAI 2016

作者： Xie, Ruobing Liu, Zhiyuan Sun, Maosong Department of Computer Science and Technology State Key Lab on Intelligent Technology and Systems National Lab for Information Science and Technology Tsinghua University Beijing China Jiangsu Collaborative Innovation Center for Language Ability Jiangsu Normal University Xuzhou221009 China

Representation learning of knowledge graphs aims to encode both entities and relations into a continuous low-dimensional vector space. Most existing methods only concentrate on learning representations with structured information located in triples, regardless of the rich information located in hierarchical types of entities, which could be collected in most knowledge graphs. In this paper, we propose a novel method named Type-embodied Knowledge Representation Learning (TKRL) to take advantages of hierarchical entity types. We suggest that entities should have multiple representations in different types. More specifically, we consider hierarchical types as projection matrices for entities, with two type encoders designed to model hierarchical structures. Meanwhile, type information is also utilized as relation-specific type constraints. We evaluate our models on two tasks including knowledge graph completion and triple classification, and further explore the performances on long-tail dataset. Experimental results show that our models significantly outperform all baselines on both tasks, especially with long-tail distribution. It indicates that our models are capable of capturing hierarchical type information which is significant when constructing representations of knowledge graphs. The source code of this paper can be obtained from https://***/thunlp/TKRL.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

Real-time Individual 3D Facial Animation by Combining Parameterized Model and Muscular Model

Real-time Individual 3D Facial Animation by Combining Parame...

引用

第35届中国控制会议

作者： Jun Yu Lingyan Li Jie Zou Department of Automation University of Science and Technology of China State Key Lab. for Novel Software Technology Nanjing University Department of Computer Science and Technology Wuhan Technology and Business University

ISBN: (纸本)9781509009107

A real-time individual 3D facial animation system is ***,facial regions and features are localized on a frontal facial image based on the combination of AdaBoost,CamShift and Active Appearance ***,an individual3 D facial mesh model,including appearance and internal articulators,is *** individual appearance mesh model is obtained by adapting an universal 3D facial mesh model to the facial features,and the individual internal articulatory mesh is obtained by integrating with the individual appearance mesh model at the same ***,to compromise between performance and realism,the parameterized model and muscular model are combined to produce facial animation based on a facial action area splitting *** results demonstrate the effectiveness and efficiency of the proposed system.

关键词： Facial model adaptation facial animation parameterized model muscular model

来源：评论

学校读者我要写书评

暂无评论

Analysis on the Content Features and Their Correlation of Web Pages for Spam Detection

引用

China Communications 2015年第3期12卷 84-94页

作者： JI Hua ZHANG Huaxiang Department of computer science Shandong Normal University Shandong Provincial key laboratory for novel distributed computer software technology

In the global information era,people acquire more and more information from the Internet,but the quality of the search results is degraded strongly because of the presence of web *** spam is one of the serious problems for search engines,and many methods have been proposed for spam *** exploit the content features of non-spam in contrast to those of *** content features for non-spam pages always possess lots of statistical regularities; but those for spam pages possess very few statistical regularities,because spam pages are made randomly in order to increase the page *** this paper,we summarize the regularities distributions of content features for non-spam pages,and propose the calculating probability formulae of the entropy and independent n-grams ***,we put forward the calculation formulae of multi features *** them,the notable content features may be used as auxiliary information for spam detection.

关键词： web spam content features feature correlation spam detection

来源：评论

学校读者我要写书评

暂无评论

A Vivid Visual Emotion Synthesis System:From Face To Hair

A Vivid Visual Emotion Synthesis System:From Face To Hair

引用

2016 IEEE 13th International Conference on Signal Processing（ICSP2016）

作者： Jun Yu Lingyan Li Department of Automation University of Science and Technology of China State Key Lab. for Novel Software Technology Nanjing University

A visual emotional synthesis system is proposed. Firstly, facial expression is synthesized by the anatomical model and parameterized model. Secondly, cartoonish hairstyle is synthesized to describe emotion implicitly by the mass-spring model and cantilever beam model. Finally, the synthesis results of facial expression and hairstyle are combined to produce a complete visual emotion synthesis result. Experiment results demonstrate the proposed system can synthesize realistic animation, and the emotion expressiveness by combining of face and hair outperform that by face or hair alone.

关键词： Visual emotion synthesis facial animation hair animation

来源：评论

学校读者我要写书评

暂无评论

Max-Margin DeepWalk: Discriminative learning of network representation 25

Max-Margin DeepWalk: Discriminative learning of network repr...

引用

25th International Joint Conference on Artificial Intelligence, IJCAI 2016

作者： Tu, Cunchao Zhang, Weicheng Liu, Zhiyuan Sun, Maosong Department of Computer Science and Technology State Key Lab on Intelligent Technology and Systems National Lab for Information Science and Technology Tsinghua University Beijing China Jiangsu Collaborative Innovation Center for Language Ability Jiangsu Normal University Xuzhou China Beijing University of Posts and Telecommunications China

DeepWalk is a typical representation learning method that learns low-dimensional representations for vertices in social networks. Similar to other network representation learning (NRL) models, it encodes the network structure into vertex representations and is learnt in unsupervised form. However, the learnt representations usually lack the ability of discrimination when applied to machine learning tasks, such as vertex classification. In this paper, we overcome this challenge by proposing a novel semi-supervised model, max-margin Deep- Walk (MMDW). MMDW is a unified NRL framework that jointly optimizes the max-margin classifier and the aimed social representation learning model. Influenced by the max-margin classifier, the learnt representations not only contain the network structure, but also have the characteristic of discrimination. The visualizations of learnt representations indicate that our model is more discriminative than unsupervised ones, and the experimental results on vertex classification demonstrate that our method achieves a significant improvement than other state-of-the-art methods. The source code can be obtained from https://github. com/thunlp/MMDW.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

Highly imbalanced classification using improved rotation forests

引用

International Journal of Wireless and Mobile Computing 2016年第1期10卷 35-41页

作者： Fang, Xiaonan Zheng, Xiyuan Tan, Yanyan Zhang, Huaxiang Department of Information Science and Engineering Shandong Provincial Key Laboratory for Novel Distributed Computer Software Technology Shandong Normal University Jinan China

Imbalanced data classification is a challenging problem in data mining. It happens in many real-world applications and has attracted growing attentions from researchers. This issue occurs when the number of one class is much higher than the other class. Ensemble of classifiers has been well known as an effective solution. Then, two novel ensemble algorithms (RUROForest and SROForest) based on rotation forests are proposed for solving highly imbalanced problems. Random undersampling or SMOTE approaches are combined with rotation forest in the proposed algorithms, which balance the uneven distribution of data sets and keep the diversity of single classifier as well. Focused on two-class highly imbalanced problems, 22 relevant data sets are performed in experiments. Experimental results and statistical analyses show that our proposed methods overtake the state-of-theart ensemble methods on the most widely used imbalanced measure criterion AUC. © Copyright 2016 Inderscience Enterprises Ltd.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

Detecting overlapping communities in networks via dominant label propagation

引用

Chinese Physics B 2015年第1期24卷 551-559页

作者：孙鹤立黄健斌田勇强宋擒豹刘怀亮 Department of Computer Science and Technology Xi’an Jiaotong University State Key Laboratory for Novel Software Technology Nanjing University School of Software Xidian University School of Economics and Management Xidian University

Community detection is an important methodology for understanding the intrinsic structure and function of a realworld network. In this paper, we propose an effective and efficient algorithm, called Dominant label Propagation Algorithm（Abbreviated as DLPA）, to detect communities in complex networks. The algorithm simulates a special voting process to detect overlapping and non-overlapping community structure in complex networks simultaneously. Our algorithm is very efficient, since its computational complexity is almost linear to the number of edges in the network. Experimental results on both real-world and synthetic networks show that our algorithm also possesses high accuracies on detecting community structure in networks.

关键词： overlapping community detection dominant label propagation complex network

来源：评论

学校读者我要写书评

暂无评论

How Effective Is Branch-Based Combinatorial Testing? An Exploratory Study

How Effective Is Branch-Based Combinatorial Testing? An Expl...

引用

IEEE International Conference on software Quality, Reliability and Security (QRS)

作者： Huiyan Wang Chang Xu Jun Sui Jian Lu Department of Computer Science and Technology Nanjing University Nanjing China State Key Laboratory for Novel Software Technology Nanjing University Nanjing China

Combinatorial testing detects faults by trying different value combinations for program inputs. Traditional combinatorial testing treats programs as black box and focuses on manipulating program inputs (named input-based combinatorial testing or ICT). In this paper, we explore the possibility of conducting combinatorial testing via white-box branch information. Similarly, different combinations of branches taken in an execution are tried to test whether they help detect faults and to what extent. We name this technique branch-based combinatorial testing (BCT). We propose ways to address challenges in realizing BCT, and evaluate BCT with Java programs. The results reported that BCT can effectively detect faults even with low-level combinations, say 3-4 ways, which suggest it to be a strong test adequacy criterion. We also found that our greedy strategy for minimizing test suites reduces over 50% tests for reaching certain way levels, and merging nested branches detects faults more cost-effectively than considering them separately.

关键词： Java software testing Fault detection Terminology software Merging

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：