检索结果-内蒙古大学图书馆

Label Informed Contrastive Pretraining for Node Importance Estimation on knowledge Graphs

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhang, Tianyu Hou, Chengbin Jiang, Rui Zhang, Xuegong Zhou, Chenghu Tang, Ke Lv, Hairong The Ministry of Education Key Laboratory of Bioinformatics Bioinformatics Division Beijing National Research Center for Information Science and Technology Department of Automation Tsinghua University Beijing100084 China The School of Computer Science and Engineering Fuyao Institute of Technology Fuzhou350109 China The State Key Laboratory of Resources and Environmental Information Systems Institute of Geographic Sciences and Natural Resources Research Chinese Academy of Sciences Beijing100101 China The Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen518055 China The Fuzhou Institute of Data Technology Fuzhou350200 China

Node Importance Estimation (NIE) is a task of inferring importance scores of the nodes in a graph. Due to the availability of richer data and knowledge, recent research interests of NIE have been dedicating to knowledge graphs for predicting future or missing node importance scores. Existing state-of-the-art NIE methods train the model by available labels, and they consider every interested node equally before training. However, the nodes with higher importance often require or receive more attention in real-world scenarios, e.g., people may care more about the movies or webpages with higher importance. To this end, we introduce Label Informed ContrAstive Pretraining (LICAP) to the NIE problem for being better aware of the nodes with high importance scores. Specifically, LICAP is a novel type of contrastive learning framework that aims to fully utilize the continuous labels to generate contrastive samples for pretraining embeddings. Considering the NIE problem, LICAP adopts a novel sampling strategy called top nodes preferred hierarchical sampling to first group all interested nodes into a top bin and a non-top bin based on node importance scores, and then divide the nodes within top bin into several finer bins also based on the scores. The contrastive samples are generated from those bins, and are then used to pretrain node embeddings of knowledge graphs via a newly proposed Predicate-aware Graph Attention Networks (PreGAT), so as to better separate the top nodes from non-top nodes, and distinguish the top nodes within top bin by keeping the relative order among finer bins. Extensive experiments demonstrate that the LICAP pretrained embeddings can further boost the performance of existing NIE methods and achieve the new state-of-the-art performance regarding both regression and ranking metrics. The source code for reproducibility is available at https://***/zhangtia16/LICAP Copyright © 2024, The Authors. All rights reserved.

关键词： knowledge graph

Effect of calcium alumina silicate hydrate nano-seeds on the hydration of low clinker cement

学校读者我要写书评

暂无评论

Journal of Building engineering 2023年 66卷

作者： Liu, Xiao Luo, Qifeng Xie, Hui Li, Shiyu Zhang, Jiale Xia, Chunlei Ding, Yungchin Chen, Yichen Gao, Ruijun Wei, Zhongyuan Zhou, Wu Wang, Ziming Cui, Suping Faculty of Materials and Manufacturing Key Laboratory of Advanced Functional Materials Ministry of Education Beijing University of Technology Beijing100124 China Faculty of Materials and Manufacturing National Engineering Laboratory for Industrial Big-data Application Technology Beijing University of Technology Beijing100124 China Beijing Municipal Engineering Research Institute Beijing100037 China Institute of Mineral Resources Engineering National Taipei University of Technology Taipei10608 China China Building Materials Academy Beijing100024 China Tangshan Longyi Tech. Co Ltd. Tangshan063004 China Guangxi Keda Building Material Chemicals Co Ltd. Nanning530221 China

The replacement of cement clinker with supplementary cementitious materials (SCMs) is considered a very effective way of carbon reduction. However, the slow pozzolanic reaction of SCMs weakens the early strength, which hinders its application. In this study, the dedicated nucleating agents for Al-rich SCMs, C-A-S-H nano-seeds, were synthesized by co-precipitation using polycarboxylate superplasticizer (PCE) as dispersant. The performances of four nano-seeds with different Al/Si and their effects on hydration in 50% metakaolin cement (MK50) were investigated in detail. The results showed that C-A-S-H nano-seeds had a greater strengthening effect than C-S-H nano-seeds especially at early ages, and also accelerated setting, with Al/Si = 0.10 being the most significant. The Al/Si = 0.10 nano-seeds advanced the second main exothermic peak by 6.7 h, promoting the production of stratlingite and causing the highest contents of CH and bound water in hydrated samples. It was also demonstrated by 27Al NMR that C-A-S-H nano-seeds enhanced the reaction degree of metakaolin by producing more C-A-S-H than AFt or AFm. The pozzolanic activity was further explored in the reaction between metakaolin and CH. This study aims to develop an efficient strengthening agent (i.e., C-A-S-H nano-seeds) for low clinker cement, and to reveal the mechanistic difference between them and common C-S-H seeds from the perspectives of hydration and pozzolanic reactions, which is of great significance for the carbon emission reduction of cement industries and solid waste utilization. © 2023 Elsevier Ltd

关键词： Hydration

On the Impact of Interlayer Misalignment for Dual-Layer data Detection in Three Dimensional Magnetic Recording

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Luo, Ke Liao, Yanzhe Zhang, Kezheng Jian, Yugen Wang, Shaobing Chen, Jincai Lu, Ping Wuhan National Laboratory for Optoelectronics Huazhong University of Science and Technology Wuhan430074 China School of Computer Science & Technology Huazhong University of Science and Technology Wuhan430074 China Key Laboratory of Information Storage System Engineering Research Center of Data Storage Systems and Technology Ministry of Education of China Huazhong University of Science and Technology Wuhan430074 China School of Public Health Hubei University of Medicine Shiyan442000 China

Three-dimensional magnetic recording (3DMR) is a crucial technology for significantly increasing the storage capacity of hard disk drives (HDDs). However, the presence of inter-symbol interference (ISI), inter-track interference (ITI), and inter-layer interference (ILI) poses significant challenges to the accurate detection of data stored on multiple layers. This study addresses the impact of interlayer misalignment on the bit error rate (BER) performance in 3DMR systems. We evaluate the performance based on a neural network estimator for reconstructing the top-layer read response signal using feedback from a Viterbi detector, enabling the separation of bottom-layer signals by subtraction from the mixed readback signal. We introduce a dual-layer partial response maximum likelihood (PRML) detector for simultaneous bit retrieval from both layers. We investigate methods of a per-layer binary classifier and a dual-layer four-class classifier based on neural networks. Our study demonstrates the BER performance of these detection schemes influenced by the interlayer misalignment, especially when the offset is 0, 10%, 50%, and 90% of the bit dimensions between the two layers. The results show that the neural network-based reconstruction and separation method achieves better bottom-layer BER performance under a slight interlayer misalignment. The BER performance benefits more from a mild downtrack offset compared with the crosstrack offset. Both the separation detection based on a neural network and the dual-layer PRML achieve the lowest top-layer BER and the worst bottom-layer BER when the interlayer misalignment is 50% of a bit size. © 2024, The Authors. All rights reserved.

关键词： Bit error rate

Ab Initio Calculation on Spectroscopic Properties and Radiative Lifetimes of Low-Lying Excited States of NaK

学校读者我要写书评

暂无评论

Chinese Journal of Chemical Physics 2019年第6期32卷 667-673,I0002页

作者： Shi-yang Zhang Feng Xie Feng-dong Jia Xiao-kang Li Ru-quan Wang Rui Li Yong Wu Zhi-ping Zhong School of Physics University of Chinese Academy of SciencesBeijing 100049China Institute of Nuclear and New Energy Technology Collaborative Innovation Center of Advanced Nuclear Energy TechnologyKey Laboratory of Advanced Reactor Engineering and Safety of Ministry of EducationTsinghua UniversityBeijing 100084China Institute of Physics Chinese Academy of SciencesBeijing 100190China Data Center for High Energy Density Physics Institute of Applied Physics and Computational MathematicsBeijing 100088China Department of Physics College of ScienceQiqihar UniversityQiqihar 161006China CAS Center for Excellence in Topological Quantum Computation University of Chinese Academy of SciencesBeijing 100190China

We performed high-level ab initio calculations on electronic structure of Na K. The potential energy curves (PECs) of 10 Λ-S states correlated with the three lowest dissociation limits have been calculated. On the basis of the calculated PECs, the spectroscopic constants of the boundΛ-S states are obtained, which are in good agreement with experimental results. The maximum vibrational quantum numbers of the singlet ground state X^1Σ^+ and the triplet ground state a^3Σ^+ have been analyzed with the semiclassical scattering theory. Transition properties including transition dipole moments, Franck-Condon factors, and radiative lifetimes have been investigated. The research results indicate that such calculations can provide fairly reliable estimation of parameters for the ultracold alkali diatomic molecular experiment.

关键词： Ultracold dipolar molecule Transition dipole moment Spectroscopic constants Potential curves

MetaGCD: Learning to Continually Learn in Generalized Category Discovery

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wu, Yanan Chi, Zhixiang Wang, Yang Feng, Songhe Key Laboratory of Big Data & Artificial Intelligence in Transportation Ministry of Education Beijing Jiaotong University Beijing100044 China School of Computer and Information Technology Beijing Jiaotong University Beijing100044 China Department of Electrical and Computer Engineering University of Toronto TorontoM5G1V7 Canada Department of Computer Science and Software Engineering Concordia University MontrealH3G2J1 Canada

In this paper, we consider a real-world scenario where a model that is trained on pre-defined classes continually encounters unlabeled data that contains both known and novel classes. The goal is to continually discover novel classes while maintaining the performance in known classes. We name the setting Continual Generalized Category Discovery (C-GCD). Existing methods for novel class discovery cannot directly handle the C-GCD setting due to some unrealistic assumptions, such as the unlabeled data only containing novel classes. Furthermore, they fail to discover novel classes in a continual fashion. In this work, we lift all these assumptions and propose an approach, called MetaGCD, to learn how to incrementally discover with less forgetting. Our proposed method uses a meta-learning framework and leverages the offline labeled data to simulate the testing incremental learning process. A meta-objective is defined to revolve around two conflicting learning objectives to achieve novel class discovery without forgetting. Furthermore, a soft neighborhood-based contrastive network is proposed to discriminate uncorrelated images while attracting correlated images. We build strong baselines and conduct extensive experiments on three widely used benchmarks to demonstrate the superiority of our method. Our code is available at https://***/ynanwu/MetaGCD. © 2023, CC0.

关键词： Learning systems

Domain-invariant Clinical Representation Learning by Bridging data Distribution Shift across EMR datasets

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Zhang, Zhongji Wang, Yuhang Zhu, Yinghao Ma, Xinyu Wang, Yasha Gao, Junyi Ma, Liantao Tang, Wen Zhang, Xiaoyun Wang, Ling Affiliated Xuzhou Municipal Hospital of Xuzhou Medical University Jiangsu China Peking University School Hospital of Stomatology Beijing China National Engineering Research Center for Software Engineering Peking University Beijing China Peking University Third Hospital Beijing China Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China Centre for Medical Informatics University of Edinburgh Edinburgh United Kingdom Health Data Research UK United Kingdom

Emerging diseases present challenges in symptom recognition and timely clinical intervention due to limited available information. An effective prognostic model could assist physicians in making accurate diagnoses and designing personalized treatment plans to prevent adverse outcomes. However, in the early stages of disease emergence, several factors hamper model development: limited data collection, insufficient clinical experience, and privacy and ethical concerns restrict data availability and complicate accurate label assignment. Furthermore, Electronic Medical Record (EMR) data from different diseases or sources often exhibit significant cross-dataset feature misalignment, severely impacting the effectiveness of deep learning models. We present a domain-invariant representation learning method that constructs a transition model between source and target datasets. By constraining the distribution shift of features generated across different domains, we capture domain-invariant features specifically relevant to downstream tasks, developing a unified domain-invariant encoder that achieves better feature representation across various task domains. Experimental results across multiple target tasks demonstrate that our proposed model surpasses competing baseline methods and achieves faster training convergence, particularly when working with limited data. Extensive experiments validate our method’s effectiveness in providing more accurate predictions for emerging pandemics and other diseases. Code is publicly available at https://***/wang1yuhang/domain_invariant_network. Copyright © 2023, The Authors. All rights reserved.

关键词： Electronic health record

Deep Domain Adaptation for Pavement Crack Detection

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Liu, Huijun Yang, Chunhua Li, Ao Huang, Sheng Feng, Xin Ruan, Zhimin Ge, Yongxin College of Computer Science Chongqing University Chongqing400044 China School of Big Data and Software Engineering Chongqing University Chongqing401331 China Chongqing City Management College Chongqing401331 China Key Laboratory of Dependable Service Computing in Cyber Physical Society Ministry of Education Chongqing University Chongqing400044 China College of Computer Science And Engineering Chongqing University of Technology Chongqing400054 China Co. Ltd Chongqing400060 China

Deep learning-based pavement cracks detection methods often require large-scale labels with detailed crack location information to learn accurate predictions. In practice, however, crack locations are very difficult to be manually annotated due to various visual patterns of pavement crack. In this paper, we propose a Deep Domain Adaptation-based Crack Detection Network (DDACDN), which learns domain invariant features by taking advantage of the source domain knowledge to predict the multi-category crack location information in the target domain, where only image-level labels are available. Specifically, DDACDN first extracts crack features from both the source and target domain by a two-branch weights-shared backbone network. And in an effort to achieve the cross-domain adaptation, an intermediate domain is constructed by aggregating the three-scale features from the feature space of each domain to adapt the crack features from the source domain to the target domain. Finally, the network involves the knowledge of both domains and is trained to recognize and localize pavement cracks. To facilitate accurate training and validation for domain adaptation, we use two challenging pavement crack datasets CQU-BPDD and RDD2020. Furthermore, we construct a new large-scale Bituminous Pavement Multi-label Disease dataset named CQU-BPMDD, which contains 38994 high-resolution pavement disease images to further evaluate the robustness of our model. Extensive experiments demonstrate that DDACDN outperforms state-of-the-art pavement crack detection methods in predicting the crack location on the target domain. Copyright © 2021, The Authors. All rights reserved.

关键词： Crack detection

Long term results in children underwent nuss procedure preschool with pectus excavatum: Real-world study with propensity matching

学校读者我要写书评

暂无评论

Heliyon 2024年第20期10卷 e39477页

作者： Wang, Quan Kong, Xiangpan Ren, Chunian Pan, Zhengxia Wu, Chun Li, Hongbo Li, Yonggang Wang, Gang Dai, Jiangtao Department of Cardiothoracic Surgery Children's Hospital of Chongqing Medical University Ministry of Education Key Laboratory of Child Development and Disorders International Science and Technology Cooperation Base of Child Development and Critical Disorders National Clinical Research Center for Child Health and Disorders Chongqing Key Laboratory of Pediatrics Chongqing Key Laboratory of Children Urogenital Development and Tissue Engineering Chongqing Higher Institution Engineering Research Center of Children's Medical Big Data Intelligent Application Chongqing China Department of Urology Children's Hospital of Chongqing Medical University China

Background: It is controversial when surgery should be performed for pectus excavatum (PE). The study retrospectively compared the long-term prognosis in PE children undergoing the Nuss procedure at the age of 4–6 versus those performing surgery at the age of 12–14. Methods: 178 patients were included in the retrospective real-world research. Clinical Characteristics were collected at baseline. The endpoint of the study was set as the last follow-up at the age of 18 in outpatient services. Satisfaction and complication rate were set as the primary outcome and were collected at the endpoint. Hospital stay post-surgery and analgesia dosage were regard as the secondary outcome. To reduce potential bias between two different age groups, propensity score matching (PSM) analysis was analysized. Results: The oral analgesic dosage of children in the 4–6 years old age group is significantly lower than that of the 12–14 years old group (0.70/kg, 0.30–1.50/kg versus 1.50/kg, 0.90–2.30/kg, P < 0.001), and children in the younger age group are discharged earlier. There was no difference in overall satisfaction (89.1 %, versus 88.20 %, P = 0.99) and recurrence rate (7.2 %, versus 6.8 %, P = 0.99) between the two groups of when they reach the age of 18. Conclusions: Performing Nuss procedure at the age of 4–6 does not affect the overall long-term satisfaction of the child or increase the long-term recurrence rate. Surgical intervention in children with pectus excavatum preschool can provide a better postoperative experience in hospital compared with intervention in adolescent. © 2024 The Authors

关键词： Long-term results Nuss procedure Pectus excavatum Preschool

Drive: An Adjustable Parallel Architecture Based on Evidence Awareness for Fake News Detection

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Cong, Mou Liu, Lu Han, Jiayu Zhu, Beibei Yuan, Xiaosong Peng, Tao College of Computer Science and Technology Jilin University 2699 Qianjin Street Jilin Changchun130012 China College of Software Jilin University 2699 Qianjin Street Jilin Changchun130012 China Key Laboratory of Symbol Computation and Knowledge Engineering Ministry of Education Jilin University 2699 Qianjin Street Jilin Changchun130012 China Department of Linguistics University of Washington SeattleWA98195 United States

Fake news detection becomes a critical challenge in NLP. Batch processing with attention-derivative models is a popular solution for fake news detection. Models with attention mechanisms could mitigate the long-distance dependence problem. However, we find that extracting local dependencies using convolutional networks is as important as extracting global dependencies using attention mechanisms. Besides, fake news datasets often suffer from problems such as imbalance and are hard to understand by current models. To address these problems, we propose DRIVE, which is short for an adjustable parallel architecture based on evidence awareness for fake news detection to alleviate the above problems, in which we introduce Claim as a supporting clue to enhance the model’s comprehension ability. Specifically, we propose a novel method for unbalanced datasets and design adjustable parallel branches, using Claim Additives Network (ClaimAddNet) and the MLP with convolutional gating mechanism (CPMLP) to capture global and local information in distinct branches, respectively. Experiments conducted on 5 public datasets demonstrate that DRIVE outperforms all baselines, improving accuracy by approximately 1.7 %. Furthermore, we implement an anti-noise interference test and evaluate the weighted average of the two branches along with ablation studies. © 2024, The Authors. All rights reserved.

关键词： Parallel architectures