检索结果-南通市图书馆

Improving Diversity with Multi-Loss Adversarial Training in Personalized News Recommendation

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第8期80卷 3107-3122页

作者： Ruijin Xue Shuang Feng qi Wang School of Computer and Cyber Sciences Communication University of ChinaBeijing100024China State Key Laboratory of Media Convergence and Communication Communication University of ChinaBeijing100024China

Users’interests are often diverse and multi-grained,with their underlying intents even more *** captur-ing users’interests and uncovering the relationships between diverse interests are key to news ***,diversity is an important metric for evaluating news recommendation algorithms,as users tend to reject excessive homogeneous information in their recommendation ***,recommendation models themselves lack diversity awareness,making it challenging to achieve a good balance between the accuracy and diversity of news *** this paper,we propose a news recommendation algorithm that achieves good performance in both accuracy and *** most existing works that solely optimize accuracy or employ more features to meet diversity,the proposed algorithm leverages the diversity-aware capability of the ***,we introduce an augmented user model to fully capture user intent and the behavioral guidance they might undergo as a ***,we focus on the relationship between the original clicked news and the augmented clicked ***,we propose an effective adversarial training method for diversity(AT4D),which is a pluggable component that can enhance both the accuracy and diversity of news recommendation *** experiments on real-world datasets confirm the efficacy of the proposed algorithm in improving both the accuracy and diversity of news recommendations.

关键词： News recommendation diversity accuracy data augmentation

Chinese Clinical Named Entity Recognition Using Multi-Feature Fusion and Multi-Scale Local Context Enhancement

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第8期80卷 2283-2299页

作者： Meijing Li Runqing Huang Xianxian qi College of Information Engineering Shanghai Maritime UniversityShanghai200306China

Chinese Clinical Named Entity Recognition(CNER)is a crucial step in extracting medical information and is of great significance in promoting medical ***,CNER poses challenges due to the specificity of clinical terminology,the complexity of Chinese text semantics,and the uncertainty of Chinese entity *** address these issues,we propose an improved CNER model,which is based on multi-feature fusion and multi-scale local context *** model simultaneously fuses multi-feature representations of pinyin,radical,Part of Speech(POS),word boundary with BERT deep contextual representations to enhance the semantic representation of text for more effective entity ***,to address the model’s limitation of focusing just on global features,we incorporate Convolutional Neural Networks(CNNs)with various kernel sizes to capture multi-scale local features of the text and enhance the model’s comprehension of the ***,we integrate the obtained global and local features,and employ multi-head attention mechanism(MHA)extraction to enhance the model’s focus on characters associated with medical entities,hence boosting the model’s *** obtained 92.74%,and 87.80%F1 scores on the two CNER benchmark datasets,CCKS2017 and CCKS2019,*** results demonstrate that our model outperforms the latest models in CNER,showcasing its outstanding overall *** can be seen that the CNER model proposed in this study has an important application value in constructing clinical medical knowledge graph and intelligent Q&A system.

关键词： CNER multi-feature fusion BiLSTM CNN MHA

Bionic immunoactivator copresenting autophagy promoting and costimulatory molecules for synergistic cancer immunotherapy

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Nano Research 2024年第3期17卷 1710-1724页

作者： Yiwei Peng Yiliang Yang Zhenzhen Yang Datong Gao Jiajia Li qi Sun Yitian Du Meng Lin Xianrong qi Beijing Key Laboratory of Molecular Pharmaceutics and New Drug Delivery System School of Pharmaceutical SciencesPeking UniversityBeijing 100191China

Immunotherapy has great promise in improving malignant tumor ***,the efficacy of existing strategies is often limited by the immunosuppressive ***,we demonstrate an in situ bionic immunoactivator,PLT-Bec1/DTA-1,with possessed natural advantages of platelets for tumor recruitment and activation,on which DTA-1(CD357 monoclonal antibody)and Bec1 were tethered as combined immune ***-Bec1/DTA-1,as a self-triggered release repository,can deliver the pre-tethered Bec1 and DTA-1 deeply through the secretion of platelet microparticles(PMPs),thereby cooperate tacitly and exhibit superiority in immune activation of dendritic cells(DCs)and T cells via autophagy inducibility,coupled with glucocorticoid-induced tumor necrosis factor receptor(GITR)-triggered T_(Reg) suppression,remodeled the immunosuppressive network of tumor ***-Bec1/DTA-1 promoted antigen presentation and T cell proliferation,and alleviated the low activity state of bone marrow-derived dendritic cells(BMDCs)in tumor suppressive ***-Bec1/DTA-1 inhibited tumor recurrence(5-and 13-fold lower of control group in tumor volume)and CD8^(+)T/T_(Reg) ratio(6.3-and 8.8-fold *** group)in mouse tumor model after intravenous or subcutaneous ***,PLT-Bec1/DTA-1 prevented tumor colonization in lung through in situ immune activation,and was slightly superior to the combined of Bec1 and *** findings highlight the promise of delivering immunostimulatory payloads via bionic carriers,eliciting automatic in situ activation of effector immune cells in tumor microenvironment for tumor *** these results provide promising prospects into the application of immunoactivator in improving cancer synergistic immunotherapy to overcome the bottlenecks in clinic.

关键词： in situ bionic immunoactivator autophagy glucocorticoid-induced tumor necrosis factor receptor(GITR)agonist platelets and platelet microparticles cancer immunotherapy

Extending the detection limit:innovations in infrared quantum dot photodetectors reaching up to 18μm

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Light(Science & Applications) 2024年第8期13卷 1445-1447页

作者： Chong Wu Wang qi Jie Wang Centre for OptoElectronics and Biophotonics School of Electrical and Electronic EngineeringNanyang Technological UniversitySingapore 639798Singapore Centre for Disruptive Photonic Technologies School of Physical and Mathematical SciencesNanyang Technological UniversitySingaporeSingapore

A regrowth method was used to synthesize large-sized colloidal quantum dots(CQDs).With the assistance of doping engineering,the synthesized CQD detectors demonstrate exceptional long-wavelength infrared detection performance,reaching up to 18μm,significantly extending the spectral response limit for CQD-based infrared *** detectors also achieve a reasonably high detectivity of 6.6×10^(8)Jones.

关键词： limit quantum sized

Boundary Points,Minimal L^(2)Integrals and Concavity PropertyⅢ—Linearity on Riemann Surfaces and Fibrations over Open Riemann Surfaces

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Acta Mathematica Sinica,English Series 2024年第9期40卷 2091-2152页

作者： qi An GUAN Zhi Tong MI Zheng YUAN School of Mathematical Sciences Peking UniversityBeijing 100871P.R.China School of Mathematics and Statistics Beijing Jiaotong UniversityBeijing 100044P.R.China Institute of Mathematics Academy of Mathematics and Systems ScienceChinese Academy of SciencesBeijing 100190P.R.China

In this article,we consider a modified version of minimal L^(2) integrals on sublevel sets of plurisubharmonic functions related to modules at boundary points,and obtain a concavity property of the modified *** applications,we give characterizations for the concavity degenerating to linearity on open Riemann surfaces and on fibrations over open Riemann surfaces.

关键词： Minimal L^(2)integrals plurisubharmonic functions boundary points weakly pseudoconvex Kähler manifold

Partition and planning:A human-like motion decision for UAV in trap environment

在线全文

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2024年第4期67卷 1226-1237页

作者： CHEN JinTao ZHOU qi REN HongRu LI HongYi School of Automation Guangdong University of TechnologyGuangzhou 510006China Provincial Key Laboratory of Intelligent Decision and Cooperative Control Guangdong University of Technology Guangzhou 510006Chi

This paper presents a human-like motion decision-making method for unmanned aerial vehicles(UAVs)navigating in trap *** proposed a space partitioning method based on sampling and consistency control to conduct a preliminary analysis of the indoor environment based on architectural *** method reduces the dimensionality of the path planning problem,thereby enhancing the ***,we designed a target-switching logic for the dynamic window *** improvement endows the UAV with the capability of both real-time obstacle avoidance and global navigation,enhancing the efficiency of the UAV in flying to task spots ***,by applying human-like methods of batch distance perception and obstacle perception to this scheme,we have further enhanced the robustness and efficiency of path ***,considering the scenario of high-rise fire rescue,we conducted simulation *** demonstrates that our scheme enhances the efficiency and robustness of path planning.

关键词： path planning indoor environment spacial partition high-rise fire rescue

Digital light processing based multimaterial 3D printing:challenges,solutions and perspectives

在线全文

学校读者我要写书评

暂无评论

International Journal of Extreme Manufacturing 2024年第4期6卷 151-174页

作者： Jianxiang Cheng Shouyi Yu Rong Wang qi Ge Shenzhen Key Laboratory of Soft Mechanics&Smart Manufacturing Southern University of Science and TechnologyShenzhen 518055People's Republic of China Department of Mechanical and Energy Engineering Southern University of Science and TechnologyShenzhen 518055People's Republic of China

Multimaterial(MM)3D printing shows great potential for application in metamaterials,flexible electronics,biomedical devices and robots,since it can seamlessly integrate distinctive materials into one printed *** numerous MM 3D printing technologies,digital light processing(DLP)MM 3D printing is compatible with a wide range of materials from hydrogels to ceramics,and can print MM 3D structures with high resolution,high complexity and fast *** paper introduces the fundamental mechanisms of DLP 3D printing,and reviews the recent advances of DLP MM 3D printing technologies with emphasis on material switching methods and material contamination *** also summarizes a number of typical examples of DLP MM 3D printing systems developed in the past decade,and introduces their system structures,working principles,material switching methods,residual resin removal methods,printing steps,as well as the representative structures and ***,we provide perspectives on the directions of the further development of DLP MM 3D printing technology.

关键词： multimaterial 3D printing digital lightprocessing multimaterial 3Dstructures

Experimental study on effects of gas flow rate on soot characteristics in diffusion flames coupled with plasma

在线全文

学校读者我要写书评

暂无评论

Science China(Technological Sciences) 2024年第2期67卷 475-485页

作者： qi DanDan CHEN MingXiao TU Xin LIU Dong MIIT Key Laboratory of Thermal Control of Electronic Equipment School of Energy and Power EngineeringNanjing University of Science and TechnologyNanjing 210094China Advanced Combustion Laboratory School of Energy and Power EngineeringNanjing University of Science and TechnologyNanjing 210094China Department of Electrical Engineering and Electronics University of LiverpoolLiverpool L693GJUK

This study examined the evolution of morphology and nanostructure of soot particles from the plasma-flame interaction for various gas flow *** current study used both optical diagnostic and sampling methods to explore the soot production and combustion *** particles were characterized at the same positions downstream from the flame zone by thermophoretic sampling and transmission electron ***,X-ray diffraction analysis,and thermogravimetric analysis were performed to study the nanostructure and oxidation reactivity of soot.A reduction in soot concentration was found with the plasma addition,which illustrated an inhibition effect of plasma on soot *** increased gas flow rate promoted soot concentration since a growing number of carbons participated in the combustion *** on the gas flow rate(carbon content)variation and plasma activation,either liquid-like soot material with irregularly shaped protrusions or chain-like structure,or a mixture of both,were generated from the diffusion *** soot produced by plasma-flame interaction also demonstrated a high correlation between nanostructure and *** soot from lower carbon content with plasma activation had a shorter fringe length and larger fringe tortuosity related to higher oxidation *** the contrary,soot from the highest carbon content without plasma-flame interaction exhibited prevalent fullerene-like nanostructures with evident large or small shells and also had a higher carbonization degree resulting in lower oxidation reactivity.

关键词： soot characteristics gas flow rate plasma diffusion flames

Improved Double Deep Q Network-Based Task Scheduling Algorithm in Edge Computing for Makespan Optimization

在线全文

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2024年第3期29卷 806-817页

作者： Lei Zeng qi Liu Shigen Shen Xiaodong Liu School of Computer Science Nanjing University of Information Science and TechnologyNanjing 210044China School of Software Nanjing University of Information Science and TechnologyNanjing 210044China School of Information Engineering Huzhou UniversityHuzhou 313000China School of Computing Edinburgh Napier UniversityEdinburghEH105DTUK

Edge computing nodes undertake an increasing number of tasks with the rise of business ***,how to efficiently allocate large-scale and dynamic workloads to edge computing resources has become a critical *** study proposes an edge task scheduling approach based on an improved Double Deep Q Network(DQN),which is adopted to separate the calculations of target Q values and the selection of the action in two networks.A new reward function is designed,and a control unit is added to the experience replay unit of the *** management of experience data are also modified to fully utilize its value and improve learning *** learning agents usually learn from an ignorant state,which is *** such,this study proposes a novel particle swarm optimization algorithm with an improved fitness function,which can generate optimal solutions for task *** optimized solutions are provided for the agent to pre-train network parameters to obtain a better cognition *** proposed algorithm is compared with six other methods in simulation *** show that the proposed algorithm outperforms other benchmark methods regarding makespan.

关键词： edge computing task scheduling reinforcement learning makespan Double Deep Q Network(DQN)