检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Xiong, Zheli Lian, Defu Chen, Enhong Chen, Gang Cheng, Xiaomin The School of Data Science University of Science and Technology of China Anhui Hefei230000 China The Anhui Province Key Laboratory of Big Data Analysis and Application School of Computer Science and Technology University of Science and Technology of China Anhui Hefei230000 China The Yangtze River Delta Information Intelligence Innovation Research Institute China

The estimation of origin-destination (OD) matrices is a crucial aspect of Intelligent Transport Systems (ITS). It involves adjusting an initial OD matrix by regressing the current observations like traffic counts of road sections (e.g., using least squares). However, the OD estimation problem lacks sufficient constraints and is mathematically underdetermined. To alleviate this problem, some researchers incorporate a prior OD matrix as a target in the regression to provide more structural constraints. However, this approach is highly dependent on the existing prior matrix, which may be outdated. Others add structural constraints through sensor data, such as vehicle trajectory and speed, which can reflect more current structural constraints in real-time. Our proposed method integrates deep learning and numerical optimization algorithms to infer matrix structure and guide numerical optimization. This approach combines the advantages of both deep learning and numerical optimization algorithms. The neural network(NN) learns to infer structural constraints from probe traffic flows, eliminating dependence on prior information and providing real-time performance. Additionally, due to the generalization capability of NN, this method is economical in engineering. We conducted tests to demonstrate the good generalization performance of our method on a large-scale synthetic dataset. Subsequently, we verified the stability of our method on real traffic data. Our experiments provided confirmation of the benefits of combining NN and numerical *** Codes I.2.1 Copyright © 2023, The Authors. All rights reserved.

关键词： Statistical tests

来源：评论

学校读者我要写书评

暂无评论

Nfgcl: A Negative-Sampling-Free Graph Contrastive Learning Framework for Recommendation

SSRN

引用

SSRN 2024年

作者： Xiao, Yuxi Ma, Rui Sang, Jun School of Big Data & Software Engineering Chongqing University Chongqing401331 China School of Computer & Network Security Chengdu University of Technology Sichuan610059 China

Recently, Contrastive Learning (CL) has seen rapid growth in graph-based recommender systems. CL-based Collaborative Filtering (CF) methods typically integrate the primary recommendation task with auxiliary CL tasks, and their notable performance can be analyzed using metrics such as uniformity and alignment in contrastive representation learning. However, most CL-based CF methods create different views of the same sample through graph augmentation in CL tasks and rely on negative sampling to distinguish user preferences. These processes are not only time-consuming but also risk learning incorrect signals. In this paper, we propose a novel Negative-sampling-Free Graph Contrastive Learning (NFGCL) framework for recommendation to balance uniformity and alignment without negative sampling. Specifically, we use in-batch positive user-item pairs to construct a preference matrix where each element represents the cosine similarity between a user and an item. We then develop a novel contrastive objective to make the preference matrix close to an identity matrix. This approach draws positive pairs closer and distances negative pairs without negative sampling, effectively improving alignment while ensuring uniformity, thereby enhancing performance. To better capture positive preferences, we design a representation-level augmentation method that integrates the output of the first-layer Graph Convolutional Networks (GCN) into the final embeddings, which improves the alignment between positive pairs. Meanwhile, we employ the uniformity loss to further regulate the uniformity of user/item representation distribution. We conduct extensive experiments on three public datasets to evaluate the effectiveness of our proposed method. The results demonstrate that our framework outperforms baseline methods and effectively balances uniformity and alignment. © 2024, The Authors. All rights reserved.

关键词： Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Robust Object Detection Based on a Comparative Learning Perspective

Robust Object Detection Based on a Comparative Learning Pers...

引用

IEEE International Symposium on Dependable, Autonomic and Secure Computing (DASC)

作者： Hao Yang Yuling Chen Hui Dou Yun Luo Chaoyue Tan Yangwen Zhang State Key Laboratory of Public Big Data College of Computer Science and Technology Guizhou University Guiyang China

With the advent of adversarial samples, the safety of Artificial Intelligence is of particular concern, as only a small amount of perturbation can be added to mislead a model's judgment. Therefore there is an urgent need for research on models that can resist adversarial perturbations. To alleviate this problem, we first analyze the vulnerability of adversarial samples and propose a uniform perspective robust model for object detection that accurately identifies both clean and ad-versarial samples. We propose a robust object detection based on the contrastive learning perspective (RCP), which can learn features from both clean and adversarial samples from a fine-grained perspective, and can recognize adversarial samples more accurately. Extensive experiments on PASCAL VOC and MS COCO show that our proposed method only degrades the clean sample detection performance by a small amount in exchange for a large robustness improvement against adversarial attacks, achieving state of the art results.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learning to Accelerate Approximate Methods for Solving Integer Programming via Early Fixing

arXiv

引用

arXiv 2022年

作者： Li, Longkang Wu, Baoyuan The School of Data Science The Chinese University of Hong Kong Shenzhen China Secure Computing Lab of Big Data Shenzhen Research Institute of Big Data Shenzhen China

Integer programming (IP) is an important but challenging problem. Approximate methods have shown promising performance on solving the IP problem. However, we observed that a large fraction of variables solved by some iterative approximate methods fluctuate around their final converged discrete states in very long iterations. It implies that these approximate methods could be significantly accelerated by early fixing these fluctuated variables to their converged states, while not significantly harming the solution quality. To this end, we propose an innovative framework of learning to early fix variables along with the approximate method. Specifically, we formulate the early fixing process as a Markov decision process, and train it using imitation learning, where a policy network evaluates the posterior probability of each free variable concerning its discrete candidate states in each block of iterations. Extensive experiments on three typical IP applications are conducted, including constrained linear programming, MRF energy minimization and sparse adversarial attack, covering moderate and large-scale IP problems. The results demonstrate that our method could not only significantly accelerate the previous approximate method up to over 10 times in most cases, but also produce similar or even better solutions. The implementation of our method is publicly available at https://***/SCLBD/Accelerated-Lpbox-ADMM. © 2022, CC BY-NC-ND.

关键词： Integer programming

来源：评论

学校读者我要写书评

暂无评论

Deep Graph-Level Clustering Using Pseudo-Label-Guided Mutual Information Maximization Network

arXiv

引用

arXiv 2023年

作者： Cai, Jinyu Han, Yi Guo, Wenzhong Fan, Jicong The College of Computer and Data Science Fuzhou University Fujian350108 China The School of Data Science The Chinese University of Hong Kong Guangdong Shenzhen518172 China The Shenzhen Research Institute of Big Data Guangdong Shenzhen518172 China

In this work, we study the problem of partitioning a set of graphs into different groups such that the graphs in the same group are similar while the graphs in different groups are dissimilar. This problem was rarely studied previously, although there have been a lot of work on node clustering and graph classification. The problem is challenging because it is difficult to measure the similarity or distance between graphs. One feasible approach is using graph kernels to compute a similarity matrix for the graphs and then performing spectral clustering, but the effectiveness of existing graph kernels in measuring the similarity between graphs is very limited. To solve the problem, we propose a novel method called Deep Graph-Level Clustering (DGLC). DGLC utilizes a graph isomorphism network to learn graph-level representations by maximizing the mutual information between the representations of entire graphs and substructures, under the regularization of a clustering module that ensures discriminative representations via pseudo labels. DGLC achieves graph-level representation learning and graph-level clustering in an end-to-end manner. The experimental results on six benchmark datasets of graphs show that our DGLC has state-of-the-art performance in comparison to many baselines. Copyright © 2023, 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

Progress and prospects of data-driven stock price forecasting research

引用

International Journal of Cognitive Computing in Engineering 2023年第1期4卷 100-108页

作者： Zhao, Chuanjun Wu, Meiling Liu, Jingfeng Duan, Zening li, Jie Shen, Lihua Shangguan, Xuekui Liu, Donghang Wang, Yanjie School of Information Shanxi University of Finance and Economics Shanxi Taiyuan030006 China School of Journalism and Mass Communication University of Wisconsin-Madison MadisonWI53706 United States Financial Innovation and Big Data Statistical Analysis Laboratory Shanxi University of Finance and Economics Shanxi Taiyuan030006 China Shanxi Information Technology Application Innovation Engineering Research Center Shanxi Taiyuan030006 China

With the rapid development of social economy and the continuous improvement of stock market, stock investment has become more and more widely concerned. Stock price prediction has become an important research direction in the field of cognitive computing in engineering. data-driven stock price forecasting aims to predict future stock price trends based on historical values and textual data, which can effectively help people reduce risks and improve returns in the process of stock investment. The article reviews the literature on stock price forecasting methods, and classifies stock price forecasting methods from two different perspectives of model and feature. According to different model angles, the existing stock price prediction methods can be divided into statistical analysis methods, traditional machine learning methods and deep learning methods. According to different characteristic angles, the existing stock price prediction methods can be divided into those based on numerical data and those based on text mixed with numerical data. Finally, we summarize the research challenges faced by stock price prediction and provide future research directions. © 2023

关键词： Investments

来源：评论

学校读者我要写书评

暂无评论

Cross-Modality Comprehensive Feature Aggregation Network for Road Scene Segmentation

Cross-Modality Comprehensive Feature Aggregation Network for...

引用

International Symposium on Computer science and Intelligent Controls (ISCSIC)

作者： Yale Yang Zihao Zhang Huifang Hou Wenwen Zhang Chunsheng Zhuang College of Information Science and Engineering Henan University of Technology Zhengzhou China Key Laboratory of Grain Information Processing and Control Ministry of Education Henan Key Laboratory of Grain Photoelectric Detection and Control and school of Artificial Intelligence and Big Data Henan University of Technology Zhengzhou China School of Artificial Intelligence and Big Data Henan University of Technology Zhengzhou China Henan Academy of Sciences Zhengzhou China

ISBN: (数字)9798350380286

ISBN: (纸本)9798350380293

RGB-thermal semantic segmentation is a potential solution for reliable semantic scene understanding under adverse weather and complex lighting conditions. However, past studies have primarily focused on the design of fusion paradigms. Unfortunately, these methods have not fully considered the modality differences caused by different imaging mechanisms, thereby reducing the performance of pixel-level semantic analysis. In response to this issue, this paper proposes an innovative Cross-Modality Comprehensive Feature Aggregation Network (CFANet). This method first designs a Cross-Modality Feature Rectification Module (CFRM) to align modality differences; then, it constructs a Global-Local Fusion Module (GLF) to sharpen object edges and refine the segmentation map. Results on two urban scene understanding datasets demonstrate that CFANet can effectively aggregate multi-layer deep features and outperform existing advanced deep learning methods in scene understanding performance.

关键词： Deep learning Semantic segmentation Roads Design methodology Semantics Lighting Imaging Reliability Meteorology Intelligent control

来源：评论

学校读者我要写书评

暂无评论

ASGEA: Exploiting Logic Rules from Align-Subgraphs for Entity Alignment

arXiv

引用

arXiv 2024年

作者： Luo, Yangyifei Chen, Zhuo Guo, Lingbing Li, Qian Zeng, Wenxuan Cai, Zhixin Li, Jianxin School of Computer Science and Engineering Beihang University China College of Computer Science and Technology Zhejiang University China Bejing Advanced Innovation Center for Big Data and Brain Computing Beihang University China

Entity alignment (EA) aims to identify entities across different knowledge graphs that represent the same real-world objects. Recent embedding-based EA methods have achieved state-of-the-art performance in EA yet faced interpretability challenges as they purely rely on the embedding distance and neglect the logic rules behind a pair of aligned entities. In this paper, we propose the Align-Subgraph Entity Alignment (ASGEA) framework to exploit logic rules from Align-Subgraphs. ASGEA uses anchor links as bridges to construct Align-Subgraphs and spreads along the paths across KGs, which distinguishes it from the embedding-based methods. Furthermore, we design an interpretable Path-based Graph Neural Network, ASGNN, to effectively identify and integrate the logic rules across KGs. We also introduce a node-level multi-modal attention mechanism coupled with multi-modal enriched anchors to augment the Align-Subgraph. Our experimental results demonstrate the superior performance of ASGEA over the existing embedding-based methods in both EA and Multi-Modal EA (MMEA) tasks. Copyright © 2024, The Authors. All rights reserved.

关键词： Computer circuits

来源：评论

学校读者我要写书评

暂无评论

INFORMATION GAIN PROPAGATION: A NEW WAY TO GRAPH ACTIVE LEARNING WITH SOFT LABELS 10

INFORMATION GAIN PROPAGATION: A NEW WAY TO GRAPH ACTIVE LEAR...

引用

10th International Conference on Learning Representations, ICLR 2022

作者： Zhang, Wentao Wang, Yexin You, Zhenbang Cao, Meng Huang, Ping Shan, Jiulong Yang, Zhi Cui, Bin School of CS Peking University China Apple National Engineering Laboratory for Big Data Analysis and Applications China Institute of Computational Social Science Peking University Qingdao China

Graph Neural Networks (GNNs) have achieved great success in various tasks, but their performance highly relies on a large number of labeled nodes, which typically requires considerable human effort. GNN-based Active Learning (AL) methods are proposed to improve the labeling efficiency by selecting the most valuable nodes to label. Existing methods assume an oracle can correctly categorize all the selected nodes and thus just focus on the node selection. However, such an exact labeling task is costly, especially when the categorization is out of the domain of individual expert (oracle). The paper goes further, presenting a soft-label approach to AL on GNNs. Our key innovations are: i) relaxed queries where a domain expert (oracle) only judges the correctness of the predicted labels (a binary question) rather than identifying the exact class (a multi-class question), and ii) new criteria of maximizing information gain propagation for active learner with relaxed queries and soft labels. Empirical studies on public datasets demonstrate that our method significantly outperforms the state-of-the-art GNN-based AL methods in terms of both accuracy and labeling cost. © 2022 ICLR 2022 - 10th International Conference on Learning Representationss. 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

OpenTM: An Open-source, Single-GPU, Large-scale Thermal Microstructure Design Framework

arXiv

引用

arXiv 2024年

作者： Quan, Yuchen Zhai, Xiaoya Fu, Xiao-Ming School of Artificial Intelligence and Data Science University of Science and Technology of China Hefei China School of Mathematical Sciences University of Science and Technology of China Hefei China

Thermal microstructures are artificially engineered materials designed to manipulate and control heat flow in unconventional ways. This paper presents an educational framework, called OpenTM, to use a single GPU for designing periodic 3D high-resolution thermal microstructures to match the predefined thermal conductivity matrices with volume fraction constraints. Specifically, we use adaptive volume fraction to make the Optimality Criteria (OC) method run stably to obtain the thermal microstructures without a large memory overhead. Practical examples with a high resolution 128 × 128 × 128 run under 90 seconds per structure on an NVIDIA GeForce GTX 4070Ti GPU with a peak GPU memory of 355 MB. Our open-source, high-performance implementation is publicly accessible at https://***/quanyuchen2000/OPENTM, and it is easy to install using Anaconda. Moreover, we provide a Python interface to make OpenTM well-suited for novices in C/C++. Copyright © 2024, The Authors. All rights reserved.

关键词： Microstructure

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：