检索结果-内蒙古大学图书馆

Better Pseudo-label: Joint Domain-aware Label and Dual-classifier for Semi-supervised Domain Generalization

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Wang, Ruiqi Qi, Lei Shi, Yinghuan Gao, Yang State Key Laboratory for Novel Software Technology Nanjing University Nanjing China National Institute of Healthcare Data Science Nanjing University Nanjing China School of Computer Science and Engineering Key Lab of Computer Network and Information Integration Ministry of Education Southeast University Nanjing China

With the goal of directly generalizing trained model to unseen target domains, domain generalization (DG), a newly proposed learning paradigm, has attracted considerable attention. Previous DG models usually require a sufficient quantity of annotated samples from observed source domains during training. In this paper, we relax this requirement about full annotation and investigate semisupervised domain generalization (SSDG) where only one source domain is fully annotated along with the other domains totally unlabeled in the training process. With the challenges of tackling the domain gap between observed source domains and predicting unseen target domains, we propose a novel deep framework via joint domain-aware labels and dual-classifier to produce high-quality pseudo-labels. Concretely, to predict accurate pseudo-labels under domain shift, a domain-aware pseudo-labeling module is developed. Also, considering inconsistent goals between generalization and pseudo-labeling: former prevents overfitting on all source domains while latter might overfit the unlabeled source domains for high accuracy, we employ a dual-classifier to independently perform pseudo-labeling and domain generalization in the training process. When accurate pseudo-labels are generated for unlabeled source domains, the domain mixup operation is applied to augment new domains between labeled and unlabeled domains, which is beneficial for boosting the generalization capability of the model. Extensive results on publicly available DG benchmark datasets show the efficacy of our proposed SSDG method. Copyright © 2021, The Authors. All rights reserved.

关键词： Image recognition

Research of Association Rules Mining based on Fuzzy Alarm Extraction

学校读者我要写书评

暂无评论

Research of Association Rules Mining based on Fuzzy Alarm Ex...

作者： Li tongyan Lei dong Qin Zhiguang Chengdu University of Information Technology Dept.of Communication Engineering Network and Data Security Key Laboratory of Sichuan Province University of Electronic Science and Technology of China Chengdu University of Information Technology Dept.of Communication Engineering

In communication alarm correlation analysis, original alarm data often have problems about information redundancy, data incompleteness, time synchronization and "sharp boundary". This paper proposed a novel alarm extraction algorithm in order to resolve above problems. We first fuzzy the original alarms to avoid "sharp boundary" during the alarm preprocessing process, and then we proposed a novel double constraint based sliding time window method to improve the efficiency on extracting data. Performance tests of the algorithm indicate that this method has higher extracting efficiency than traditional methods, which is suitable for alarm data in telecommunication networks.

关键词： Alarm correlation analysis fuzzy alarm sharp boundary

Erratum to “ECMER: Edge-Cloud Collaborative Personalized Multimodal Emotion Recognition Framework in the Internet of Vehicles”

学校读者我要写书评

暂无评论

IEEE network 2024年第3期38卷 292-292页

作者： Puning Zhang Miao Fu Rongjian Zhao Dapeng Wu Hongbin Zhang Zhigang Yang Ruyan Wang School of Communication and Information Engineering Chongqing University of Posts and Telecommunications Chongqing China Advanced Network and Intelligent Connection Technology Key Laboratory of Chongqing Education Commission of China and the Chongqing Key Laboratory of Ubiquitous Sensing and Networking Chongqing China Chongqing Innovation Center of Industrial Big-Data Company Ltd. Chongqing China

This addresses errors in [1] . There is incorrect author affiliation information in this article.

关键词： Emotion recognition Collaboration Internet of Vehicles

Label-consistency based graph neural networks for semi-supervised node classification

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Xu, Bingbing Huang, Junjie Hou, Liang Shen, Huawei Gao, Jinhua Cheng, Xueqi CAS Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences China School of Computer and Control Engineering University of Chinese Academy of Sciences China

Graph neural networks (GNNs) achieve remarkable success in graph-based semi-supervised node classification, leveraging the information from neighboring nodes to improve the representation learning of target node. The success of GNNs at node classification depends on the assumption that connected nodes tend to have the same label. However, such an assumption does not always work, limiting the performance of GNNs at node classification. In this paper, we propose label-consistency based graph neural network (LC-GNN), leveraging node pairs unconnected but with the same labels to enlarge the receptive field of nodes in GNNs. Experiments on benchmark datasets demonstrate the proposed LC-GNN outperforms traditional GNNs in graph-based semi-supervised node classification. We further show the superiority of LC-GNN in sparse scenarios with only a handful of labeled nodes. Copyright © 2020, The Authors. All rights reserved.

关键词： Graph neural networks

Graph wavelet neural network 7

学校读者我要写书评

暂无评论

Graph wavelet neural network

7th International Conference on Learning Representations, ICLR 2019

作者： Xu, Bingbing Shen, Huawei Cao, Qi Qiu, Yunqi Cheng, Xueqi CAS Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences China School of Computer and Control Engineering University of Chinese Academy of Sciences Beijing China

We present graph wavelet neural network (GWNN), a novel graph convolutional neural network (CNN), leveraging graph wavelet transform to address the shortcomings of previous spectral graph CNN methods that depend on graph Fourier transform. Different from graph Fourier transform, graph wavelet transform can be obtained via a fast algorithm without requiring matrix eigendecomposition with high computational cost. Moreover, graph wavelets are sparse and localized in vertex domain, offering high efficiency and good interpretability for graph convolution. The proposed GWNN significantly outperforms previous spectral graph CNNs in the task of graph-based semi-supervised classification on three benchmark datasets: Cora, Citeseer and Pubmed. © 7th International Conference on Learning Representations, ICLR 2019. All Rights Reserved.

关键词： Wavelet transforms

Privacy-preserving data outsourcing with integrity auditing for lightweight devices in cloud computing 14th

学校读者我要写书评

暂无评论

Privacy-preserving data outsourcing with integrity auditing ...

14th International Conference on Information Security and Cryptology, Inscrypt 2018

作者： Liu, Dengzhi Shen, Jian Chen, Yuling Wang, Chen Zhou, Tianqi Wang, Anxi Jiangsu Engineering Center of Network Monitoring Nanjing University of Information Science and Technology Nanjing China Guizhou Provincial Key Laboratory of Public Big Data Guiyang China

ISBN: (纸本)9783030142339

The cloud can provide unlimited storage space to users via the Internet. Unlike locally data storing, users will lose the direct control of the data after outsourcing it to the cloud. Moreover, the cloud is an untrusted entity. It is possible that the cloud may try to extract, discard and destroy users’ data due to benefits. Hence, the data security in cloud computing needs to be well guaranteed. In this paper, we propose a privacy-preserving data outsourcing scheme with integrity auditing for lightweight devices in cloud computing. On the one hand, the blind signature is used in the proposed scheme to delegate the generation of users’ data signatures to the TPA. On the other hand, based on the property of the BLS signature, the blinded signatures received from the TPA can be verified by the user and the data integrity stored in the cloud can be audited by the TPA. In addition, the proposed scheme supports batch operation. Security analysis shows that the proposed scheme achieves the properties of correctness, privacy-preserving and non-forgeability. Performance analysis indicates that the proposed scheme can be performed with high efficiency. © 2019, Springer Nature Switzerland AG.

关键词： Outsourcing

AI exposure predicts unemployment risk

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Frank, Morgan R. Ahn, Yong-Yeol Moro, Esteban Department of Informatics and Networked Systems University of Pittsburgh PittsburghPA15216 United States Digital Economy Lab Institute for Human-Centered Artificial Intelligence Stanford University StanfordCA94305 United States Media Laboratory Massachusetts Institute of Technology CambridgeMA02139 United States Connection Science Massachusetts Institute of Technology CambridgeMA United States Center for Complex Networks and Systems Research Luddy School of Informatics Computing and Engineering Indiana University BloomingtonIN United States Indiana University Network Science Institute Indiana University BloomingtonIN United States Department of Mathematics & GISC Universidad Carlos III de Madrid Leganes28911 Spain Institute for Data Systems and Society Massachusetts Institute of Technology CambridgeMA United States

Is artificial intelligence (AI) disrupting jobs and creating unemployment? Despite many attempts to quantify occupations' exposure to AI, inconsistent validation obfuscates the relative benefits of each approach. A lack of disaggregated labor outcome data, including unemployment data, further exacerbates the issue. Here, we assess which models of AI exposure predict job separations and unemployment risk using new occupation-level unemployment data by occupation from each US state's unemployment insurance office spanning 2010 through 2020. Although these AI exposure scores have been used by governments and industry, we find that individual AI exposure models are not predictive of unemployment rates, unemployment risk, or job separation rates. However, an ensemble of those models exhibits substantial predictive power suggesting that competing models may capture different aspects of AI exposure that collectively account for AI's variable impact across occupations, regions, and time. Our results also call for dynamic, context-aware, and validated methods for assessing AI exposure. Interactive visualizations for this study are available at https://***/~mrfrank/uiRiskDemo/. © 2023, CC BY.

关键词： Visualization

Repeated-root constacyclic codes with optimal locality

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Zhao, Wei Shum, Kenneth W. Yang, Shenghao The School of Science and Engineering The Chinese University of Hong Kong Shenzhen China University of Science and Technology of China China The School of Science and Engineering The Chinese University of Hong Kong Guangdong Shenzhen518172 China The Shenzhen Key Laboratory of IoT Intelligent Systems and Wireless Network Technology The Chinese University of Hong Kong Guangdong Shenzhen518172 China The Shenzhen Research Institute of Big Data Guangdong Shenzhen518172 China

A code is called a locally repairable code (LRC) if any code symbol is a function of a small fraction of other code symbols. When a locally repairable code is employed in a distributed storage systems, an erased symbol can be recovered by accessing only a small number of other symbols, and hence alleviating the network resources required during the repair process. In this paper we consider repeated-root constacyclic codes, which is a generalization of cyclic codes, that are optimal with respect to a Singleton-like bound on minimum distance. An LRC with the structure of a constacyclic code can be encoded efficiently using any encoding algorithm for constacyclic codes in general. In this paper we obtain optimal LRCs among these repeated-root constacyclic codes. Several infinite classes of optimal LRCs over a fixed alphabet are found. Under a further assumption that the ambient space of the repeated-root constacyclic codes is a chain ring, we show that there is no other optimal LRC. © 2021, CC BY.

关键词： Multiprocessing systems

Boosting Connectivity in Retinal Vessel Segmentation via a Recursive Semantics-Guided network 23rd

学校读者我要写书评

暂无评论

Boosting Connectivity in Retinal Vessel Segmentation via a R...

23rd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2020

作者： Xu, Rui Liu, Tiantian Ye, Xinchen Lin, Lin Chen, Yen-Wei DUT-RU International School of Information Science and Engineering Dalian University of Technology Dalian China DUT-RU Co-Research Center of Advanced ICT for Active Life Dalian China Key Laboratory for Ubiquitous Network and Service Software of Liaoning Province Dalian China College of Software Dalian University of Technology Dalian China College of Information Science and Engineering Ritsumeikan University Kusatsu Japan Research Center of Healthcare Data Science Zhejiang Lab Hangzhou China

ISBN: (纸本)9783030597214

Many deep learning based methods have been proposed for retinal vessel segmentation, however few of them focus on the connectivity of segmented vessels, which is quite important for a practical computer-aided diagnosis system on retinal images. In this paper, we propose an efficient network to address this problem. A U-shape network is enhanced by introducing a semantics-guided module, which integrates the enriched semantics information to shallow layers for guiding the network to explore more powerful features. Besides, a recursive refinement iteratively applies the same network over the previous segmentation results for progressively boosting the performance while increasing no extra network parameters. The carefully designed recursive semantics-guided network has been extensively evaluated on several public datasets. Experimental results have shown the efficiency of the proposed method. © 2020, Springer Nature Switzerland AG.

关键词： Semantics