检索结果-内蒙古大学图书馆

A Knowledge-Injected Curriculum Pretraining Framework for Question Answering

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Lin, Xin Su, Tianhuang Huang, Zhenya Xue, Shangzi Liu, Haifeng Chen, Enhong School of Computer Science and Technology University of Science and Technology of China State Key Laboratory of Cognitive Intelligence Hefei China Guangdong OPPO Mobile Telecommunications Corp. Ltd Shenzhen China University of Science and Technology of China Hefei China Anhui Province Key Laboratory of Big Data Analysis and Application University of Science and Technology of China State Key Laboratory of Cognitive Intelligence Hefei China

Knowledge-based question answering (KBQA) is a key task in natural language processing research, and also an approach to access the web data and knowledge, which requires exploiting knowledge graphs (KGs) for reasoning. In the literature, one promising solution for KBQA is to incorporate the pretrained language model (LM) with KGs by generating KG-centered pretraining corpus, which has shown its superiority. However, these methods often depend on specific techniques and resources to work, which may not always be available and restrict its application. Moreover, existing methods focus more on improving language understanding with KGs, while neglect the more important human-like complex reasoning. To this end, in this paper, we propose a general Knowledge-Injected Curriculum Pretraining framework (KICP) to achieve comprehensive KG learning and exploitation for KBQA tasks, which is composed of knowledge injection (KI), knowledge adaptation (KA) and curriculum reasoning (CR). Specifically, the KI module first injects knowledge into the LM by generating KG-centered pretraining corpus, and generalizes the process into three key steps that could work with different implementations for flexible application. Next, the KA module learns knowledge from the generated corpus with LM equipped with an adapter as well as keeps its original natural language understanding ability to reduce the negative impacts of the difference between the generated and natural corpus. Last, to enable the LM with complex reasoning, the CR module follows human reasoning patterns to construct three corpora with increasing difficulties of reasoning, and further trains the LM from easy to hard in a curriculum manner to promote model learning. We provide an implementation of the general framework, and evaluate the proposed KICP on four real-word datasets. The results demonstrate that our framework can achieve higher performances, and have good generalization ability to other QA tasks. Copyright © 2024, The

关键词： Knowledge graph

Source-free domain adaptation based on label reliability for cross-domain bearing fault diagnosis

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Wu, Wenyi Zhang, Hao Wei, Zhisen Jing, Xiao-Yuan Zhang, Qinghua Wu, Songsong Guangdong Provincial Key Laboratory of Petrochemical Equipment Fault Diagnosis School of Computer Guangdong University of Petrochemical Technology Maoming525000 China Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen518055 China Key Laboratory of Data Science and Intelligence Application Minnan Normal University Fujian Province Zhangzhou363000 China

Source-free domain adaptation (SFDA) has been exploited for cross-domain bearing fault diagnosis without access to source data. Current methods select partial target samples with reliable pseudo-labels for model adaptation, which is sub-optimal due to the ignored target samples. We argue that every target sample can contribute to model adaptation, and accordingly propose in this paper a novel SFDA-based approach for bearing fault diagnosis that exploits both reliable and unreliable pseudo-labels. We develop a data-augmentation-based label voting strategy to divide the target samples into reliable and unreliable ones. We propose to explore the underlying relation between feature space and label space by using the reliable pseudo-labels as ground-truth labels, meanwhile, alleviating negative transfer by maximizing the entropy of the unreliable pseudo-labels. The proposed method achieves well-balance between discriminability and diversity by taking advantage of reliable and unreliable pseudo-labels. Extensive experiments are conducted on two bearing fault benchmarks, demonstrating that our approach achieves significant performance improvements against existing SFDA-based bearing fault diagnosis methods. Our code is available at https://***/BdLab405/SDALR Copyright © 2025, The Authors. All rights reserved.

关键词： Failure analysis

Optimal analysis of penalized lowest-order mixed FEMs for the Stokes–Darcy model

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Cao, Luling Sun, Weiwei School of Mathematics and Data Science Shaanxi University of Science and Technology Xi’an710021 China Advanced Institute of Natural Science Beijing Normal University Zhuhai China Guangdong Provincial Key Laboratory of Interdisciplinary Research and Application for Data Science BNU-HKBU United International College Zhuhai519087 China

This paper is concerned with non-uniform fully-mixed FEMs for dynamic coupled Stokes–Darcy model with the well-known Beavers–Joseph–Saffman (BJS) interface condition. In particular, a decoupled algorithm with the lowest-order mixed non-uniform FE approximations (MINI for the Stokes equation and RT0-DG0 for the Darcy equation) and the classical Nitsche-type penalty is studied. The method with the combined approximation of different orders is commonly used in practical simulations. However, the optimal error analysis of methods with non-uniform approximations for the coupled Stokes–Darcy flow model has remained challenging, although the analysis for uniform approximations has been well done. The key question is how the lower-order approximation to the Darcy flow influences the accuracy of the Stokes solution through the interface condition. In this paper, we prove that the decoupled algorithm provides a truly optimal convergence rate in L2-norm in spatial direction: O(h2) for Stokes velocity and O(h) for Darcy flow in the coupled Stokes–Darcy model. This implies that the lower-order approximation to the Darcy flow does not pollute the accuracy of numerical velocity for Stokes flow. The analysis presented in this paper is based on a well-designed Stokes–Darcy Ritz projection and given for a dynamic coupled model. The optimal error estimate holds for more general combined approximations and more general coupled models, including the corresponding model of steady-state Stokes–Darcy flows and the model of coupled dynamic Stokes and steady-state Darcy flows. Numerical results confirm our theoretical analysis and show that the decoupled algorithm is efficient. Copyright © 2025, The Authors. All rights reserved.

关键词： Convergence of numerical methods

Joint Timing and Frequency Synchronization for Coherent Optical SEFDM Systems

学校读者我要写书评

暂无评论

Joint Timing and Frequency Synchronization for Coherent Opti...

2023 Asia Communications and Photonics Conference/2023 International Photonics and Optoelectronics Meetings, ACP/POEM 2023

作者： Shi, Jinze Du, Xinwei Liu, Shuai Wang, Qian Yu, Changyuan BNU-HKBU United International College Guangdong Provincial Key Laboratory of Interdisciplinary Research and Application for Data Science Zhuhai519087 China College of Information Engineering Zhejiang University of Technology Hangzhou China Photonics Research Institute The Hong Kong Polytechnic University Department of Electronic and Information Engineering Hung Hom Hong Kong

ISBN: (纸本)9798350312614

In this paper, we propose a joint timing and frequency synchronization algorithm for spectrally efficient frequency-division multiplexing (SEFDM) systems by utilizing a training symbol composed of a conjugate symmetric sequence. The timing metric is designed to be the multiplication of two timing metrics which aims to reduce the amplitude of sidelobes, thus ensuring the timing synchronization performance. In addition, according to the phase relationship between the four halves in the designed training symbol, we then derive a closed form expression of the the carrier frequency offset (CFO) estimate. Simulation results verify the superior timing and frequency synchronization accuracy of the proposed method compared with the existing synchronization algorithms. © 2023 IEEE.

关键词： carrier frequency offset Spectrally efficient frequency-division multiplexing timing synchronization

Transmitter IQ Mismatch Estimation by Number-Theoretic Net-Based Gaussian Particle Filter for Coherent Optical Communications

学校读者我要写书评

暂无评论

Transmitter IQ Mismatch Estimation by Number-Theoretic Net-B...

2023 Asia Communications and Photonics Conference/2023 International Photonics and Optoelectronics Meetings, ACP/POEM 2023

作者： Liu, Shuai Xu, Yangfan Du, Xinwei Wang, Qian Yu, Changyuan BNU-HKBU United International College Guangdong Provincial Key Laboratory of Interdisciplinary Research and Application for Data Science Zhuhai519087 China College of Information Engineering Zhejiang University of Technology Hangzhou China Photonics Research Institute The Hong Kong Polytechnic University Department of Electronic and Information Engineering Hung Hom Hong Kong

ISBN: (纸本)9798350312614

In this paper, we propose a novel approach for estimating in-phase/quadrature-phase (IQ) mismatch at the transmitter using a number-theoretic net (NT-net)-based Gaussian particle filter (NT-GPF). The proposed method generates particles on an ellipse of the given bivariate Gaussian distribution, which helps locate the correct region of the actual state parameters more efficiently. Simulation results demonstrate that the NT-GPF-based approach offers higher estimation accuracy, with the mean squared error (MSE) consistently lower than that of traditional GPF across different particle numbers. Moreover, the proposed NT-GPF-based algorithm significantly improves estimation efficiency and noise tolerance, requiring only 21 particles under low signal-to-noise ratio (SNR). © 2023 IEEE.

关键词： Gaussian particle filter IQ mismatch number-theoretic net

Pilot-Aided Deep Learning Based Phase Estimation for OFDM Systems with Wiener Phase Noise

学校读者我要写书评

暂无评论

Pilot-Aided Deep Learning Based Phase Estimation for OFDM Sy...

2023 Asia Communications and Photonics Conference/2023 International Photonics and Optoelectronics Meetings, ACP/POEM 2023

作者： Wang, Qian Chen, Xingke Qian, Liping Du, Xinwei Yu, Changyuan Kam, Pooi-Yuen College of Information Engineering Zhejiang University of Technology Hangzhou310023 China BNU-HKBU United International College Guangdong Provincial Key Laboratory of Interdisciplinary Research and Application for Data Science Zhuhai519087 China The Hong Kong Polytechnic University Department of Electronic and Information Engineering 999077 Hong Kong Shenzhen518172 China

ISBN: (纸本)9798350312614

Orthogonal frequency-division multiplexing (OFDM) and high-order modulations have a wide range of applications in coherent optical communications. However, the existence of phase noise will greatly affect the system performance. To overcome this issue, this paper proposes a pilot-aided deep learning (PADL)-based phase estimation scheme, since deep learning has been a hot trend to be applied on digital signal processing in communications. To be specific, preliminary phase noise is first estimated by the pilot-aided (PA) method, and then the estimates are fed into the neural network to train for more accurate estimation. Simulation results show that the mean square error performance of the proposed method is much better than the conventional PA method. Especially for high-order modulations (M>64), the bit error rate of the PADL-based receiver is smaller than that with even double pilots used for phase estimation, which verifies a stronger robustness in our design using limited spectrum resources. © 2023 IEEE.

关键词： deep learning estimation OFDM phase noise

Optimal Trajectory Planning and Task Assignment for UAV-assisted Fog Computing 24

学校读者我要写书评

暂无评论

Optimal Trajectory Planning and Task Assignment for UAV-assi...

24th IEEE International Conference on High Performance Computing and Communications, 8th IEEE International Conference on data science and Systems, 20th IEEE International Conference on Smart City and 8th IEEE International Conference on Dependability in Sensor, Cloud and Big data Systems and application, HPCC/DSS/SmartCity/DependSys 2022

作者： Liu, Shuaijun Yin, Jiaying Zeng, Zishu Wu, Jingjin BNU-HKBU United International College Department of Statistics and Data Science Zhuhai519087 China The University of Hong Kong Li Ka Shing Faculty of Medicine Pokfulam Hong Kong School of Engineering The Hong Kong University of Science and Technology Kowloon Hong Kong Guangdong Provincial Key Laboratory of Interdisciplinary Research and Application for Data Science China

ISBN: (纸本)9798350319934

Fog computing is an emerging distributed computing model for the realization of the Internet of Things (IoT). It extends computing and caching functions to the edge of wireless networks. Unmanned Aerial Vehicles (UAVs) provide effective support for fog computing, as UAVs can not only act as a relay between mobile users and physically remote edge devices to avoid costly long-range wireless communications, but also are equipped with computing facilities that can take over certain tasks. In this paper, we aim to optimize the energy efficiency of a fog computing system assisted by a single UAV, by planning the trajectories of the UAV and assigning computing tasks to different devices including the UAV itself. We propose two algorithms based on the classical Ant Colony and Particle Swarm Optimization techniques, and solve the problem by continuous convex approximation. Different from most existing studies where the trajectories are assumed to be straight lines, we account for the effect of obstacles, such as buildings, and deliberately avoid them during the trajectory planning phase. We demonstrate, through extensive simulation experiments, that our proposed approach can achieve significantly better energy efficiency compared to existing benchmark algorithms. © 2022 IEEE.

关键词： Fog computing

Grazing exclusion promotes soil organic carbon accumulation in Tibetan grasslands with lower temperatures

学校读者我要写书评

暂无评论

Ecological Processes 2024年第4期13卷 254-264页

作者： Guangru Zhang Xiangping Tan Jinhong He Dengnan Luo Xiang Zeng Minqi Liang Ruochen Cao Siyuan Peng Pan Li Long Tao Muhammed Mustapha Ibrahim Zhongmin Hu School of Geography South China Normal UniversityGuangzhou 510631China Guangdong Provincial Key Laboratory of Applied Botany South China Botanical GardenChinese Academy of Sciences723 Xingke Rd.Tianhe DistrictGuangzhou 510650China Guangzhou Institute of Forestry and Landscape Architecture Guangzhou Urban Ecosystem National Field StationGuangzhou 510405China Facility of Geographical Science Beijing Normal UniversityBeijing 100875China College of Arts and Sciences Beijing Normal University at ZhuhaiZhuhai 519087China School of Atmospheric Sciences Guangdong Province Data Center of Terrestrial and Marine Ecosystems Carbon CycleSun Yat-Sen UniversityZhuhai 510245China International Institute for Earth System Sciences Jiangsu Center for Collaborative Innovation in Geographical Information Resource Development and ApplicationNanjing UniversityNanjing 210023China Jiangsu Provincial Key Laboratory of Geographic Information Science and Technology Key Laboratory for Land Satellite Remote Sensing Applications of Ministry of Natural ResourcesSchool of Geography and Ocean ScienceNanjing UniversityNanjing 210023China Hainan Baoting Tropical Rainforest Ecosystem Observation and Research Station School of EcologyHainan University58 Renmin AvenueHaikou 570228China

Background Grazing exclusion is a practical approach to restore vegetation in degraded grasslands and enhance soil organic carbon(SOC)***,the dynamics and drivers of SOC in grasslands after grazing exclusion have not been well documented,especially in ecosystems with cold *** Here,we established 14 paired treatments(grazing exclusion ***-grazing)along a 600-km transect in the northeastern zone of the Qinghai-Tibet *** six years,we analyzed vegetation biomass dynamics and measured the soil physicochemical properties and organic C concentration across three depths(0–10,10–20,and 20–30 cm).Results Grazing exclusion signifcantly increased above-and belowground biomass(139.85%and 43.30%,respectively),pH(1.38%),total phosphorus(3.29%),nitrate nitrogen(18.03%),and ammonium nitrogen(17.81%),but signifcantly decreased soil bulk density(2.43%)and clay content(10.49%),particularly in 0–30 ***,SOC concentrations positively responded to grazing exclusion(0–10 cm)in 9 of the 14 sites *** efects of grazing exclusion on SOC concentrations were signifcantly higher in areas with a mean annual temperature(MAT)below 0℃ compared to those in sites with a high MAT(>0℃).The SOC concentrations signifcantly correlated with the mean annual precipitation(MAP)in both treatments,but these correlations diminished with increasing soil *** regression analysis showed that soil chemical properties(e.g.,total nitrogen and phosphorus)positively infuenced SOC accumulation,while MAT negatively infuenced it after grazing *** analysis further revealed that MAT indirectly regulated SOC dynamics via soil chemical *** Our study highlights that grazing exclusion results in an asynchronous SOC and plant biomass accumulation and may be more benefcial for SOC sequestration in Qinghai-Tibet Plateau grasslands with lower ***,humid climates promote SOC concentration in alpine *** results

关键词： Grazing exclusion Plant biomass Qinghai-Tibet Plateau Soil carbon storage Soil physicochemical property

Prediction of vaccination coverage level in the heterogeneous mixing population

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Bai, Fan Chen, Qianyu Xu, Yizhuo Department of Mathematical Sciences Faculty of Science and Technology Beijing Normal University Hong Kong Baptist University United International College Guangdong Zhuhai China Guangdong Provincial Key Laboratory of Interdisciplinary Research and Application for Data Science BNU-HKBU United International College Guangdong Zhuhai China

Heterogeneity of population is a key factor in modeling the transmission of disease among the population and has huge impact on the outcome of the transmission. In order to investigate the decision making process in the heterogeneous mixing population regarding whether to be vaccinated or not, we propose the modeling framework which includes the epidemic models and the game theoretical analysis. We consider two sources of heterogeneity in this paper: the different activity levels and the different relative vaccination costs. It is interesting to observe that, if both sources of heterogeneity are considered, there exist a finite number of Nash equilibria (evolutionary stable strategies (ESS)) of the vaccination game. While if only the difference of activity levels is considered, there are infinitely many Nash equilibira. For the latter case, the initial condition of the decision making process becomes highly sensitive. In the application of public health management, the inclusion of population heterogeneity significantly complicates the prediction of the overall vaccine coverage level. © 2024, CC BY.

关键词： Transmissions