检索结果-南通市图书馆

在线全文

学校读者我要写书评

暂无评论

基于自注意力机制的图卷积协同过滤模型

合肥学院学报（综合版） 2023年第2期40卷 74-80,89页

作者：孙克雷孙文鑫安徽理工大学计算机科学与工程学院安徽淮南232000

针对目前已有的基于图学习的推荐系统(Graph Learning Based Recomender Systems,GLRS)存在的过平滑问题,提出了一种基于图注意力机制的协同过滤模型,对于传统的基于图卷积的协同过滤模型(Neural Graph Collaborative Filtering,NGCF),... 详细信息

针对目前已有的基于图学习的推荐系统(Graph Learning Based Recomender Systems,GLRS)存在的过平滑问题,提出了一种基于图注意力机制的协同过滤模型,对于传统的基于图卷积的协同过滤模型(Neural Graph Collaborative Filtering,NGCF),当堆叠多层时,结点嵌入变得更加相似,最终难以区分,直接影响到推荐性能。本文认为NGCF中,没有相似兴趣的高阶邻居用户也等权重的参与了在图卷积操作中的用户嵌入的学习,因此想到在消息传播过程中融合了图注意力机制计算邻居权重,以此来取代现有方法中通过哈达玛积的形式来学习用户和物品间的交互编码。本文在三个公共数据集上做了实验,其NDCG@20和Recall在Gowalla数据集上分别提升了1.943%和2.129%。

关键词：推荐系统图卷积网络自注意力机制协同过滤

AI for Supporting the Freedom of Drawing

在线全文

学校读者我要写书评

暂无评论

Machine Intelligence Research 2024年第1期21卷 63-88页

作者： Xiaohua sun Juexiao Qin College of Design and Innovation Tongji UniversityShanghai 200082China

Artificial intelligence(AI)has recently been developing rapidly in image processing and *** is not only starting to take over repetitive and tedious tasks but is also involved in creative *** is one of the areas with the greatest potential for collaboration between humans and *** is an important approach to express various information,but due to the lack of appropriate knowledge and skills,ordinary people without long-time training are unable to draw *** there have been various attempts at human-AI collaboration in drawing,it is difficult for researchers to consider the wide variety of specific problems and develop univer-sal methods due to the openness,improvisation,and individuality of *** this paper,we first analysed the contents of drawing and the general creation process in ***,we have described a mechanism for using AI to enable people to regain the freedom of drawing ***,we have developed a framework that describes methods for analysing specific problems and quickly finding solutions by building connections between the influencing factors in drawing,the demands of humans,and possible implementa-tion *** framework also reveals a broad scope of possibilities for applying AI to support people in drawing.

关键词： Intention understanding drawing support drawing art human-artificial intelligence(AI)collaboration

Catalysis by subsurface metal atoms in Al(001) through electron transfer

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

National Science Review 2024年第4期11卷 12-13页

作者： Bo sun Ding Ma Beijing National Laboratory for Molecular Science New Cornerstone Science Laboratory College of Chemistry and Molecular Engineering Peking University

Heterogeneous catalysis plays a crucial role in the chemical industry, which has evolved from the investigation of chemisorption to an electronic structure[1]. Diverse strategies, such as alloying,doping, and confining, have been applied to enhance the catalytic properties of active sites through electronic modulation. The electron transfer between active sites and substrates profoundly influences catalytic activity [2].

关键词：

Persistent Variations in the East Asian Trough from March to April and the Possible Mechanism

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Advances in Atmospheric Sciences 2024年第4期41卷 737-753,I0002-I0004页

作者： Shui YU Jianqi sun Nansen-Zhu International Research Center Institute of Atmospheric PhysicsChinese Academy of SciencesBeijing 100029China College of Earth and Planetary Sciences University of Chinese Academy of SciencesBeijing 100049China

The East Asian trough(EAT)profoundly influences the East Asian spring *** this study,the relationship of the EATs among the three spring months is *** analysis shows that the variation in March EAT is closely related to that of April *** empirical orthogonal function(EEOF)analysis also confirms the co-variation of the March and April *** positive/negative EEOF1 features the persistent strengthened/weakened EAT from March to *** investigation indicates that the variations in EEOF1 are related to a dipole sea surface temperature(SST)pattern over the North Atlantic and the SST anomaly over the tropical Indian *** dipole SST pattern over the North Atlantic,with one center east of Newfoundland Island and another east of Bermuda,could trigger a Rossby wave train to influence the EAT in March−*** SST anomaly over the tropical Indian Ocean can change the Walker circulation and influence the atmospheric circulation over the tropical western Pacific,subsequently impacting the southern part of the EAT in March−*** the SST factors,the Northeast Asian snow cover could change the regional thermal conditions and lead to persistent EAT anomalies from March to *** three impact factors are generally independent of each other,jointly explaining large variations in the EAT ***,the signals of the three factors could be traced back to February,consequently providing a potential prediction source for the EAT variation in March and April.

关键词： spring East Asian trough sea surface temperature snow cover

Pretreating and normalizing metabolomics data for statistical analysis

在线全文

学校读者我要写书评

暂无评论

Genes & Diseases 2024年第3期11卷 188-205页

作者： Jun sun Yinglin Xia Division of Gastroenterology and Hepatology Department of MedicineDepartment of Microbiology/ImmunologyUIC Cancer CenterUniversity of Illinois ChicagoJesse Brown VA Medical Center Chicago(537)ChicagoIL 60612USA Division of Gastroenterology and Hepatology Department of MedicineUniversity of Illinois ChicagoChicagoIL 60612USA

Metabolomics as a research field and a set of techniques is to study the entire small molecules in biological *** is emerging as a powerful tool generally for pre-cision ***,integration of microbiome and metabolome has revealed the mechanism and functionality of microbiome in human health and ***,metabo-lomics data are very ***/pretreating and normalizing procedures on metabolomics data are usually required before statistical *** this review article,we comprehensively review various methods that are used to preprocess and pretreat metabolo-mics data,including MS-based data and NMR-based data preprocessing,dealing with zero and/or missing values and detecting outliers,data normalization,data centering and scaling,data *** discuss the advantages and limitations of each *** choice for a suitable preprocessing method is determined by the biological hypothesis,the characteristics of the data set,and the selected statistical data analysis *** then provide the perspective of their applications in the microbiome and metabolome research.

关键词： Data centering and scaling Data normalization Data transformation Missing values MS-Baseddata preprocessing NMRData preprocessing Outliers Preprocessing/pretreatment

In silico optimization of actuation performance in dielectric elastomercomposites via integrated finite element modeling and deep learning

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Theoretical & Applied Mechanics Letters 2024年第1期14卷 48-56页

作者： Jiaxuan Ma Sheng sun Materials Genome Institute Shanghai UniversityShanghai 200444China Zhejiang Laboratory Hangzhou 311100China Shanghai Frontier Science Center of Mechanoinformatics Shanghai UniversityShanghai 200444China

Dielectric elastomers(DEs)require balanced electric actuation performance and mechanical integrity under applied *** high dielectric particles as fillers provides extensive design space to optimize concentration,morphology,and distribution for improved actuation performance and material *** study presents an integrated framework combining finite element modeling(FEM)and deep learning to optimize the microstructure of DE *** first calculates actuation performance and the effective modulus across varied filler combinations,with these data used to train a convolutional neural network(CNN).Integrating the CNN into a multi-objective genetic algorithm generates designs with enhanced actuation performance and material modulus compared to the conventional optimization approach based on FEM approach within the same *** framework harnesses artificial intelligence to navigate vast design possibilities,enabling optimized microstructures for high-performance DE composites.

关键词： Dielectric elastomer composites Multi-objective optimization Finite element modeling Convolutional neural network

Improved Twin Support Vector Machine Algorithm and Applications in Classification Problems

在线全文

学校读者我要写书评

暂无评论

China Communications 2024年第5期21卷 261-279页

作者： sun Yi Wang Zhouyang Beijing University of Posts and Telecommunications School of Computer Science(National Pilot Software Engineering School) Beijing University of Posts and Telecommunications International School

The distribution of data has a significant impact on the results of *** the distribution of one class is insignificant compared to the distribution of another class,data imbalance *** will result in rising outlier values and ***,the speed and performance of classification could be greatly *** the above problems,this paper starts with the motivation and mathematical representing of classification,puts forward a new classification method based on the relationship between different classification *** with the vector characteristics of the actual problem and the choice of matrix characteristics,we firstly analyze the orderly regression to introduce slack variables to solve the constraint problem of the lone *** we introduce the fuzzy factors to solve the problem of the gap between the isolated points on the basis of the support vector *** introduce the cost control to solve the problem of sample ***,based on the bi-boundary support vector machine,a twostep weight setting twin classifier is *** can help to identify multitasks with feature-selected patterns without the need for additional optimizers,which solves the problem of large-scale classification that can’t deal effectively with the very low category distribution gap.

关键词： fuzzy ordered regression(OR) relaxing variables twin support vector machine

A Method for Detecting and Recognizing Yi Character Based on Deep Learning

在线全文

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第2期78卷 2721-2739页

作者： Haipeng sun Xueyan Ding Jian sun HuaYu Jianxin Zhang School of Computer Science and Engineering Dalian Minzu UniversityDalian116600China Institute of Machine Intelligence and Bio-Computing Dalian Minzu UniversityDalian116600China 3Yi Language Research RoomChina Ethnic Languages Translation Centre Beijing100080China

Aiming at the challenges associated with the absence of a labeled dataset for Yi characters and the complexity of Yi character detection and recognition,we present a deep learning-based approach for Yi character detection and *** the detection stage,an improved Differentiable Binarization Network(DBNet)framework is introduced to detect Yi characters,in which the Omni-dimensional Dynamic Convolution(ODConv)is combined with the ResNet-18 feature extraction module to obtain multi-dimensional complementary features,thereby improving the accuracy of Yi character ***,the feature pyramid network fusion module is used to further extract Yi character image features,improving target recognition at different ***,the previously generated feature map is passed through a head network to produce two maps:a probability map and an adaptive threshold map of the same size as the original *** maps are then subjected to a differentiable binarization process,resulting in an approximate binarization *** map helps to identify the boundaries of the text ***,the text detection box is generated after the post-processing *** the recognition stage,an improved lightweight MobileNetV3 framework is used to recognize the detect character regions,where the original Squeeze-and-Excitation(SE)block is replaced by the efficient Shuffle Attention(SA)that integrates spatial and channel attention,improving the accuracy of Yi characters ***,the use of depth separable convolution and reversible residual structure can reduce the number of parameters and computation of the model,so that the model can better understand the contextual information and improve the accuracy of text *** experimental results illustrate that the proposed method achieves good results in detecting and recognizing Yi characters,with detection and recognition accuracy rates of 97.5%and 96.8%,*** also,we have compared the detection and re

关键词： Yi characters text detection text recognition attention mechanism deep neural network

The Dbar-dressing method for the(2+1)-dimensional Date–Jimbo–Kashiwara–Miwa equation

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Communications in Theoretical Physics 2024年第1期76卷 17-23页

作者： Shifei sun Biao Li School of Mathematics and Statistics Ningbo UniversityNingbo315211China

In this work,the(2+1)-dimensional Date–Jimbo–Kashiwara–Miwa(DJKM)equation is studied by means of the ■-dressing method.A new ■ problem has been constructed by analyzing the characteristic function and the Green’s function of its Lax *** on solving the ■ equation and choosing the proper spectral transformation,the solution of the DJKM equation is ***,the more general solution of the DJKM equation can be also obtained by ensuring the evolution of the time spectral data.

关键词： (2+1)-dimensional DJKM equation inverse spectral problem Green's function characteristic function ■-dressing method