检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Xu, Zhuo Bai, Lu Cui, Lixin Li, Ming Wang, Yue Hancock, Edwin R. School of Artificial Intelligence Beijing Normal University Beijing China School of Information Central University of Finance and Economics Beijing China The Key Laboratory of Intelligent Education Technology Application of Zhejiang Province Zhejiang Normal University Jinhua China Department of Computer Science University of York York United Kingdom

Graph Auto-Encoders (GAEs) are powerful tools for graph representation learning. In this paper, we develop a novel Hierarchical Cluster-based GAE (HC-GAE), that can learn effective structural characteristics for graph data analysis. To this end, during the encoding process, we commence by utilizing the hard node assignment to decompose a sample graph into a family of separated subgraphs. We compress each subgraph into a coarsened node, transforming the original graph into a coarsened graph. On the other hand, during the decoding process, we adopt the soft node assignment to reconstruct the original graph structure by expanding the coarsened nodes. By hierarchically performing the above compressing procedure during the decoding process as well as the expanding procedure during the decoding process, the proposed HC-GAE can effectively extract bidirectionally hierarchical structural features of the original sample graph. Furthermore, we re-design the loss function that can integrate the information from either the encoder or the decoder. Since the associated graph convolution operation of the proposed HC-GAE is restricted in each individual separated subgraph and cannot propagate the node information between different subgraphs, the proposed HC-GAE can significantly reduce the over-smoothing problem arising in the classical convolution-based GAEs. The proposed HC-GAE can generate effective representations for either node classification or graph classification, and the experiments demonstrate the effectiveness on real-world datasets. Copyright © 2024, The Authors. All rights reserved.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Bone metastasis segmentation based on Improved U-NET algorithm 4

Bone metastasis segmentation based on Improved U-NET algorit...

引用

2021 4th International Conference on Advanced Algorithms and Control Engineering, ICAACE 2021

作者： Zhang, Jingyi Huang, Mengge Deng, Tao Cao, Yongchun Lin, Qiang Key Laboratory of China's Ethnic Languages and Information Technology of Ministry of Education Northwest Minzu University Lanzhou730030 China Key Laboratory of Streaming Data Computing Technologies and Application Northwest Minzu University Lanzhou730030 China School of Mathematics and Computer Science Northwest Minzu University Lanzhou730030 China

Whole body bone scan image analysis is widely used in nuclear medicine to assist nuclear medicine physicians in the detection of bone metastases. At present, the analysis of whole-body bone scan images mainly relies on the manual reading of nuclear medicine doctors. The doctors, based on personal knowledge and experience, look for abnormal lesion locations and diagnose them by examining the whole-body bone scan images. However, this method is prone to misdiagnosis and missed diagnosis. To solve the above problems, this study proposes an image segmentation method based on deep learning, which can automatically identify the location of bone metastases, so that doctors can make more accurate diagnosis. The Methods Attention mechanism was added to the jump connection of the original U-NET network to enhance the image feature selection. Experiments show that the algorithm in this study teaches traditional U-Net to show better results on the three indicators of MIoU Dice and MAP. © Published under licence by IOP Publishing Ltd.

关键词： Diagnosis

来源：评论

学校读者我要写书评

暂无评论

ENADPool: The Edge-Node Attention-based Differentiable Pooling for Graph Neural Networks

arXiv

引用

arXiv 2024年

作者： Zhao, Zhehan Bai, Lu Cui, Lixin Li, Ming Wang, Yue Xu, Lixiang Hancock, Edwin R. School of Artificial Intelligence Beijing Normal University Beijing China School of Information Central University of Finance and Economics Beijing China The Key Laboratory of Intelligent Education Technology and Application of Zhejiang Province Zhejiang Normal University Jinhua China Department of Computer Science University of York York United Kingdom

Graph Neural Networks (GNNs) are powerful tools for graph classification. One important operation for GNNs is the downsampling or pooling that can learn effective embeddings from the node representations. In this paper, we propose a new hierarchical pooling operation, namely the Edge-Node Attention-based Differentiable Pooling (ENADPool), for GNNs to learn effective graph representations. Unlike the classical hierarchical pooling operation that is based on the unclear node assignment and simply computes the averaged feature over the nodes of each cluster, the proposed ENADPool not only employs a hard clustering strategy to assign each node into an unique cluster, but also compress the node features as well as their edge connectivity strengths into the resulting hierarchical structure based on the attention mechanism after each pooling step. As a result, the proposed ENADPool simultaneously identifies the importance of different nodes within each separated cluster and edges between corresponding clusters, that significantly addresses the shortcomings of the uniform edge-node based structure information aggregation arising in the classical hierarchical pooling operation. Moreover, to mitigate the over-smoothing problem arising in existing GNNs, we propose a Multi-distance GNN (MD-GNN) model associated with the proposed ENADPool operation, allowing the nodes to actively and directly receive the feature information from neighbors at different random walk steps. Experiments demonstrate the effectiveness of the MD-GNN associated with the proposed ENADPool. Copyright © 2024, The Authors. All rights reserved.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Structured information extraction for bone scan image report texts 4

Structured information extraction for bone scan image report...

引用

2021 4th International Conference on Advanced Algorithms and Control Engineering, ICAACE 2021

作者： Huang, Mengge Zhang, Jingyi Deng, Tao Man, Zhengxing Lin, Qiang Key Laboratory of China's Ethnic Languages and Information Technology of Ministry of Education Northwest Minzu University Lanzhou730030 China Key Laboratory of Streaming Data Computing Technologies and Application Northwest Minzu University Lanzhou730030 China School of Mathematics and Computer Science Northwest Minzu University Lanzhou730030 China

Extracting structured information from the bone scan image report text plays a crucial role in supporting clinical analysis and research. This study summarized the structure and characteristics of 3608 bone scan image report text using dictionary-based information extraction method, including data cleaning, entity recognition, building dictionary and extraction rules. This method was used to obtain the structured data of bone scan image report text required for clinical research, and the effect evaluation was carried out on 1000 randomly selected report texts, with the precision rate and recall rate higher than 90%. The method proposed in this study is practical and could have good effect on structured results for bone scan imaging report text. © Published under licence by IOP Publishing Ltd.

关键词： Clinical research

来源：评论

学校读者我要写书评

暂无评论

Electrical Transport Properties of Few-Layer SnS2 Field-effect Transistors

Electrical Transport Properties of Few-Layer SnS2 Field-effe...

引用

2022 International Conference on Electrical, Electronics and Information Engineering, EEIE 2022

作者： Zhang, Lu Li, Tuo Yao, Peng Zou, Xiaofeng Ji, Chao Zhan, Yongzheng He, Jiaqi Zhang, Qing State Key Laboratory of High-end Server & Storage Technology Jinan250013 China Inspur Electronic Information Industry Co. Ltd. Jinan250013 China Shandong Yunhai Guochuang Cloud Computing Equipment Industry Innovation Co. Ltd. Jinan250013 China Beijing Institute of Computer Technology and Application Beijing100854 China College of Mathematics and Physics Beijing University of Chemical Technology Beijing100029 China

After the discovery of graphene in 2004, two dimensional (2D) materials have fascinated a lot of view due to the excellent properties. Nowadays, the research on 2D materials has spread to other graphene-like layer structured materials, especially transition metal dichalcogenides (TMDCs). Tin disulfide (SnS2) is a kind of TMDCs with a sizable bandgap. Here we introduce few-layer SnS2 field-effect transistors (FETs) fabricated using micromechanical exfoliation method. The FETs show n-type behavior, the on/off ratio exceeding 0.54×104 and the carrier mobility is 0.61 cm2V-1s-1. The electronic and optical characteristics of SnS2 flakes with a finite bandgap illustrate their potential applications in optoelectronics device. © Published under licence by IOP Publishing Ltd.

关键词： Field effect transistors

来源：评论

学校读者我要写书评

暂无评论

Collaborative filtering recommendation algorithm integrating user similarity and trust 4

Collaborative filtering recommendation algorithm integrating...

引用

2021 4th International Conference on computer Information Science and application technology, CISAT 2021

作者： Huang, Mengge Cao, Kai Zhang, Jingyi Zhang, Chuanlin Deng, Tao Key Laboratory of China's Ethnic Languages and Information Technology of Ministry of Education Northwest Minzu University Lanzhou730030 China Key Laboratory of Streaming Data Computing Technologies and Application Northwest Minzu University Lanzhou730030 China School of Mathematics and Computer Science Northwest Minzu University Lanzhou730030 China

The rapid development of the Internet has brought convenience to people and has also produced the problem of "information overload". In view of the traditional collaborative filtering algorithm facing some bottlenecks to be solved, this study proposes a collaborative filtering algorithm that combines similarity and trust. First of all, in view of the large deviation of traditional similarity calculation and prediction of user ratings, this study proposes an optimized Pearson correlation coefficient calculation method;secondly, the trust relationship is established based on the user's rating of the common project, and the trust relationship between users who do not have a direct trust relationship is established through the transfer of trust;then find the nearest neighbor set of the target user through the fusion of user similarity and trust;finally, the item is scored and predicted to generate a recommendation list. Experimental results show that the algorithm proposed in this study can effectively improve the accuracy of recommendation. © Content from this work may be used under the terms of the Creative Commons Attribution 3.0 Licence.

关键词： Collaborative filtering

来源：评论

学校读者我要写书评

暂无评论

Deep learning-based multi-modal data integration enhancing breast cancer disease-free survival prediction

引用

Precision Clinical Medicine 2024年第2期7卷 132-145页

作者： Zehua Wang Ruichong Lin Yanchun Li Jin Zeng Yongjian Chen Wenhao Ouyang Han Li Xueyan Jia Zijia Lai Yunfang Yu Herui Yao Weifeng Su Guangdong Key Laboratory of Cross-Application of Data Science and Technology Beijing Normal University-Hong Kong Baptist University United International CollegeZhuhai 519087China Faculty of Innovation Engineering Macao University of Science and TechnologyTaipaMacao 999078China Department of Computer and Information Engineering Guangzhou Huali CollegeGuangzhou 511325China Department of Pathology Sun Yat-sen Memorial HospitalSun Yat-sen UniversityGuangzhou 510120China Guangzhou National Laboratory Guangzhou 510005China Dermatology and Venereology Division Department of Medicine SolnaCenter for Molecular MedicineKarolinska InstitutetStockholm 17177Sweden Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation Department of Medical OncologyBreast Tumor CentrePhase I Clinical Trial CentreSun Yat-sen Memorial HospitalSun Yat-sen UniversityGuangzhou 510120China The Second Clinical Medical College Southern Medical UniversityGuangzhou 510515China Faculty of Medicine Macao University of Science and TechnologyTaipaMacao 999078China

Background:The prognosis of breast cancer is often unfavorable,emphasizing the need for early metastasis risk detection and accurate treatment *** study aimed to develop a novel multi-modal deep learning model using preoperative data to predict disease-free survival(DFS).Methods:We retrospectively collected pathology imaging,molecular and clinical data from The Cancer Genome Atlas and one independent institution in *** developed a novel Deep Learning Clinical Medicine Based Pathological Gene Multi-modal(DeepClinMed-PGM)model for DFS prediction,integrating clinicopathological data with molecular *** patients included the training cohort(n=741),internal validation cohort(n=184),and external testing cohort(n=95).Result:Integrating multi-modal data into the DeepClinMed-PGM model significantly improved area under the receiver operating characteristic curve(AUC)*** the training cohort,AUC values for 1-,3-,and 5-year DFS predictions increased to 0.979,0.957,and 0.871,while in the external testing cohort,the values reached 0.851,0.878,and 0.938 for 1-,2-,and 3-year DFS predictions,*** DeepClinMed-PGM's robust discriminative capabilities were consistently evident across various cohorts,including the training cohort[hazard ratio(HR)0.027,95%confidence interval(CI)0.0016-0.046,P<0.0001],the internal validation cohort(HR 0.117,95%CI 0.041-0.334,P<0.0001),and the external cohort(HR 0.061,95%CI 0.017-0.218,P<0.0001).Additionally,the DeepClinMed-PGM model demonstrated C-index values of 0.925,0.823,and 0.864 within the three cohorts,***:This study introduces an approach to breast cancer prognosis,integrating imaging and molecular and clinical data for enhanced predictive accuracy,offering promise for personalized treatment strategies.

关键词： breast cancer multi-modality deep learning pathological disease-free survival

来源：评论

学校读者我要写书评

暂无评论

Reinforcement Learning-Based Explainable Recommendation over Knowledge Graphs with Negative Sampling

Reinforcement Learning-Based Explainable Recommendation over...

引用

Symposia and Workshops on Ubiquitous, Autonomic and Trusted Computing, UIC-ATC

作者： Siyuan Zhang Yuanxin Ouyang Zhuang Liu Wenge Rong Zhang Xiong State Key Laboratory of Software Development Environment Beihang University Beijing China Engineering Research Center of Advanced Computer Application Technology Ministry of Education Beihang University Beijing China

Introducing knowledge graphs (KGs) into the recommender systems not only improves their performance but also enhances the interpretability. However, most KG-based recommendation methods have the problem of inefficiency and ex-post explanation, which reinforcement learning (RL) methods can solve properly. Most existing RL-based methods for explainable recommendations only consider positive rewards when designing the reward part of the RL environment, which is defective and misleads the policy of the RL agent. To address this problem, we propose Reinforced Knowledge Graph Reasoning with Reinforced Negative Sampling (RKGR-RNS) by introducing a negative sampling method into RL-based recommendation, which refines the reward mechanism to help optimize the agent’s policy. And a judge module is proposed to improve the performance of the recommender system further. Experiments on three real datasets demonstrate that our method is better than the state-of-the-art baseline.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multimodal Remote Sensing Image Registration Based on Local Phase Congruency and keypoint Filtering

Multimodal Remote Sensing Image Registration Based on Local ...

引用

Chinese Automation Congress (CAC)

作者： Tao Ma Weijian Huang Linlin Du Chun Wang Shuangquan Ge Yunfei Liu Kun Yu Institute of Computer Application China Academy of Engineering Physics Mianyang China National Key Laboratory of Science and Technology on Vessel Integrated Power System Naval University of Engineering Wuhan China

Robust registration is able to register the complementary information captured by multimodal images. Owning to the nonlinear intensity differences and repeating patterns in multimodal images, many classic descriptors are unable to achieve good performance. Aiming at this problem, a remote sensing image registration method based on key point description and filtering is proposed. First, this work utilizes a new keypoint descriptor with Log-Gabor filters to characterize the oriented phase congruency information in multimodal images. Then, a keypoint filtering is performed on all descriptors by mean-shift clustering to remove the repeating patterns. This approach is evaluated on three commonly used data sets, which are used for structure map construction and image registration comparisons. The experimental results show that the proposed method is able to preserve highly similar structure features and remove most repeating keypoints in multimodal images, which indicates this method is efficient and can achieve good registration performance.

关键词： Image registration Automation Information filters Feature extraction Computational efficiency Registers Remote sensing

来源：评论

学校读者我要写书评

暂无评论

MAKT: Multichannel Attention Networks based Knowledge Tracing with Representation Learning

MAKT: Multichannel Attention Networks based Knowledge Tracin...

引用

IEEE International Conference on Teaching, Assessment and Learning for Engineering (TALE)

作者： Xuyang Jiang Yuanxin Ouyang Zhuang Liu Wenge Rong Zhang Xiong State Key Laboratory of Software Development Environment Beihang University Beijing China Engineering Research Center of Advanced Computer Application Technology Ministry of Education Beihang University Beijing China

As an effective and emerging component of intelligent education, Knowledge Tracing(KT) achieves the combination of artificial intelligence and individualized learning, whose aim is to assess students’ mastery of knowledge concepts and assist in developing learning plans. Several existing KT models either use concepts sequence as input and evaluate students’ knowledge state or treat exercise as input to predict students’ future performance. In this paper, we introduce a constraint factor to extract concepts’ and exercises’ relation matrix, design three methods in representation learning, and propose a Multichannel Attention Networks based KT model(MAKT). Specifically, we restrict the co-occurrence relationship within a time window to extract the relation matrix and then train their representations via graph generative learning, graph contrastive learning, and matrix decomposition, respectively. In MAKT, a sliding window is implemented by multichannel where input sequence is sequentially lagged in turn by one position and attention mechanism is applied. We conduct experiments on several benchmark datasets and demonstrate that MAKT with concepts’ and exercises’ representation trained by matrix decomposition outperforms state-of-the-art models.

关键词：

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：