检索结果-内蒙古大学图书馆

12th International Conference on Learning Representations, ICLR 2024

作者： Zhang, Zihan Lee, Jason D. Chen, Yuxin Du, Simon S. Department of Electrical and Computer Engineering Princeton University United States Department of Statistics and Data Science University of Pennsylvania United States Paul G. Allen School of Computer Science and Engineering University of Washington United States

A recent line of works showed regret bounds in reinforcement learning (RL) can be (nearly) independent of planning horizon, a.k.a. the horizon-free bounds. However, these regret bounds only apply to settings where a polynomial dependency on the size of transition model is allowed, such as tabular Markov Decision Process (MDP) and linear mixture MDP. We give the first horizon-free bound for the popular linear MDP setting where the size of the transition model can be exponentially large or even uncountable. In contrast to prior works which explicitly estimate the transition model and compute the inhomogeneous value functions at different time steps, we directly estimate the value functions and confidence sets. We obtain the horizon-free bound by: (1) maintaining multiple weighted least square estimators for the value functions;and (2) a structural lemma which shows the maximal total variation of the inhomogeneous value functions is bounded by a polynomial factor of the feature dimension. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved.

关键词： Markov processes

来源：评论

学校读者我要写书评

暂无评论

Uncertainty propagation in the internet of things

引用

Discover Internet of Things 2024年第1期4卷 1-20页

作者： Pal, Shantanu Khalifa, Sara Miller, Dimity Dedeoglu, Volkan Dorri, Ali Ramachandran, Gowri Moghadam, Peyman Kusy, Brano Jurdak, Raja The School of Information Technology Faculty of Science Engineering and Built Environment Deakin University MelbourneVIC 3125 Australia The School of Information Systems Faculty of Science Queensland University of Technology BrisbaneQLD4000 Australia The School of Electrical Engineering and Robotics Faculty of Engineering Queensland University of Technology BrisbaneQLD4000 Australia The Distributed Sensing Systems Group Data61 CSIRO PullenvaleQLD 4069 Australia The School of Computer Science Faculty of Science Queensland University of Technology BrisbaneQLD 4000 Australia The CSIRO Robotics Data61 CSIRO PullenvaleQLD 4069 Australia

The Internet of Things (IoT) detects context through sensors capturing data from dynamic physical environments, in order to inform automation decisions within cyber physical systems (CPS). Diverse types of uncertainty in the IoT pipeline can propagate within and across nodes, involving complex interactions with security, privacy, and trust that remain largely unexplored. This paper conducts an in-depth analysis of the types of uncertainties in IoT and how they propagate within IoT nodes and networks. We consider adversarial uncertainty in the context of distributed IoT networks to capture perturbations due to malicious actors in the network that can influence IoT security, privacy and trust. We examine the propagation of adversarial uncertainty and well-known uncertainty types, namely aleatoric and epistemic uncertainty, within the five distinct stages of sensing, communication, storage, processing and decision-making in an IoT pipeline, across network layer boundaries, and across nodes within an IoT network. Using this mapping, we analyse the interactions between the uncertainty types and their propagation, and security, privacy, and trust in IoT. Based on this analysis, we discuss guidelines and considerations for mitigating uncertainty in IoT through a smart grid use case study. © The Author(s) 2024.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

An improved branching algorithm for the proper interval edge deletion problem

引用

Frontiers of computer science 2022年第2期16卷 178-180页

作者： Wenjun LI Xiaojing TANG Yongjie YANG School of Computer and Communication Engineering Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on TransportationChangsha University of Science and TechnologyChangsha 410015China School of Computer Science and Engineering Central South UniversityChangsha 410083China

1 Introduction For a graph class G,the G-EDGE DELETION problem is to determine whether a given graph can be transformed into a graph in G by deleting at most k *** G-EDGE DELETION problem for a large body of graph cla... 详细信息

关键词： branching problem transformed

来源：评论

学校读者我要写书评

暂无评论

Scalable Algorithms for Densest Subgraph Discovery 39

Scalable Algorithms for Densest Subgraph Discovery

引用

39th IEEE International Conference on data engineering, ICDE 2023

作者： Luo, Wensheng Tang, Zhuo Fang, Yixiang Ma, Chenhao Zhou, Xu Hunan University College of Computer Science and Electronic Engineering China The Chinese University of HongKong School of Data Science Shenzhen China

ISBN: (纸本)9798350322279

As a fundamental problem in graph data mining, Densest Subgraph Discovery (DSD) aims to find the subgraph with the highest density from a graph. It has been studied for several decades and found a large number of real-world applications, such as network community detection, regulatory motif discovery in DNA, graph index construction, and fake follower detection. Although there are many existing DSD algorithms, they are often not scalable or efficient to process large-scale graphs, since most of them are serial algorithms and can only leverage the computing resource of a single CPU core. To tackle these issues, in this paper we propose efficient parallel algorithms for solving the DSD problems on both undirected and directed graphs at scale. Our main idea is to use the k-cores (a kind of dense subgraph) to approximate the densest subgraph in the undirected graphs, and then propose efficient parallel algorithms for computing the cores by optimizing the iterative process and also reducing the number of iterations. We further extend this idea for directed graphs by introducing a novel concept, named w-induced subgraph, to avoid unnecessary enumerations of x or y when searching [x,y]-cores (a kind of directed dense subgraph to approximate the densest). To verify the scalability and efficiency of the proposed algorithms, we have conducted extensive experiments on 12 large real-world graphs, and four of them are billion-scale. The experimental results show that our proposed algorithms outperform the state-of-the-art algorithms on both undirected and directed graphs, in terms of scalability and efficiency. © 2023 IEEE.

关键词： Directed graphs

来源：评论

学校读者我要写书评

暂无评论

Deep mining algorithm for traffic accident data on highways under severe weather conditions

Advances in Transportation Studies

引用

Advances in Transportation Studies 2024年第SPEICAL ISSUE期1卷 47-58页

作者： Zhang, J.Y. Cao, Y. Jiangsu Youth Science and Technology Center Nanjing210000 China School of Computer Engineering Jiangsu Second Normal University Nanjing210000 China Jiangsu Province Engineering Research Center of Basic Education Big Data A PPlication Nanjing210013 China

Analyzing traffic accident data is crucial for pinpointing contributing factors’ forecasting accident patterns’ and informing effective safety measures. This insight leads to enhanced road safety’ decreased fatalities’ and better resource allocation in transportation planning. Severe weather conditions’ including rain’ fog’ and sandstorms’ can significantly reduce the visibility of highways and increase the risk of traffic accidents. The paper proposes a deep mining algorithm for road traffic accident data under adverse weather conditions. Firstly’ collect traffic accident data under adverse weather conditions and complete attribute reduction’ denoising’ and normalization processing. Then’ by improving the decision tree algorithm’ the attribute selection criteria of the decision tree algorithm were enhanced;Finally’ the improved decision tree algorithm is aPPlied to achieve deep mining of traffic accident data. The experimental results show that proposed algorithm has an accuracy rate of over 95.3%’ an accuracy rate of over 94.3%’ a recall rate of over 92.7%’ and a stable F1 value of around 0.8’ demonstrating excellent performance. The provided deep mining algorithm for traffic accident data can provide more scientific and accurate decision suPPort for traffic management departments. © 2024, Aracne Editrice. All rights reserved.

关键词： data mining

来源：评论

学校读者我要写书评

暂无评论

Design, Mathematical Modeling, and Control of an Underactuated 3-DOF Experimental Helicopter

引用

IEEE Access 2024年 12卷 55568-55586页

作者： Makki, Osamah Talal Moosapour, Seyyed Sajjad Mobayen, Saleh Nobari, Jafar Heyrani Shahid Chamran University of Ahvaz Faculty of Engineering Ahvaz*** Iran National Yunlin University of Science and Technology Graduate School of Intelligent Data Science Douliu Yunlin640301 Taiwan K. N. Toosi University of Technology Department of Electrical and Computer Engineering Tehran19697 Iran

This article covers the design, implementation, mathematical modelling, and control of a multivariable, underactuated, low-cost, three-degrees-of-freedom experimental helicopter system (namely a 3-DOF helicopter). The system is considered an essential benchmark for the development and evaluation of control algorithms and identification methods as it mimics real helicopters and UAVs. The system is multivariable and difficult to control due to its underactuation characteristics. In addition, it presents further challenges in modelling and control analysis in terms of high nonlinearity, coupling, uncertainty, and unmodeled dynamics. In this scene, the system is designed, developed, and operated virtually in SolidWorks using the necessary hardware and software. A new accurate multivariable coupled nonlinear mathematical model is derived and proposed. The model takes into account the complex dynamics and interactions between the system components, providing a more comprehensive representation of the real system compared to previous models. System parameters are then identified using the Sum of the Square Error (SSE) algorithm, and the mathematical model is compared to the practical results in many different scenarios via Simulink. Experimental tests confirmed the mathematical model and robust performance of the implemented experimental 3-DOF helicopter system, as well as the ability of the proposed controller to stabilize the helicopter and track different trajectories in the presence of uncertainties and cross-coupling effects. © 2013 IEEE.

关键词： Iron

来源：评论

学校读者我要写书评

暂无评论

Sketch2Human: Deep Human Generation with Disentangled Geometry and Appearance Constraints

引用

IEEE Transactions on Visualization and computer Graphics 2024年 PP卷 1-14页

作者： Qu, Linzi Shang, Jiaxiang Ye, Hui Han, Xiaoguang Fu, Hongbo School of Creative Media City University of Hong Kong Hong Kong Department of Computer Science & Engineering HKUST Hong Kong Shenzhen Research Institute of Big Data Chinese University of Hong Kong Shenzhen China

Geometry- and appearance-controlled full-body human image generation is an interesting but challenging task. Existing solutions are either unconditional or dependent on coarse conditions (e.g., pose, text), thus lacking explicit geometry and appearance control of body and garment. Sketching offers such editing ability and has been adopted in various sketch-based face generation and editing solutions. However, directly adapting sketch-based face generation to full-body generation often fails to produce high-fidelity and diverse results due to the high complexity and diversity in the pose, body shape, and garment shape and texture. Recent geometrically controllable diffusion-based methods mainly rely on prompts to generate appearance. It is hard to balance the realism and the faithfulness of their results to the sketch when the input is coarse. This work presents Sketch2Human, the first system for controllable full-body human image generation guided by a semantic sketch (for geometry control) and a reference image (for appearance control). Our solution is based on the latent space of StyleGAN-Human with inverted geometry and appearance latent codes as input. Specifically, we present a sketch encoder trained with a large synthetic dataset sampled from StyleGAN-Human's latent space and directly supervised by sketches rather than real images. Considering the entangled information of partial geometry and texture in StyleGAN-Human and the absence of disentangled datasets, we design a novel training scheme that creates geometry-preserved and appearance-transferred training data to tune a generator to achieve disentangled geometry and appearance control. Although our method is trained with synthetic data, it can also handle hand-drawn sketches. Qualitative and quantitative evaluations demonstrate the superior performance of our method to state-of-the-art methods. IEEE

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Multicomponent Similarity Graphs for Cross-Network Node Classification

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第3期5卷 1411-1424页

作者： Zhang, Yuhong Shi, Congmei Li, Xinzheng Zhang, Zan Hu, Xuegang School of Computer Science and Information Engineering Hefei University of Technology Hefei230601 China Institute of Artificial Intelligence Hefei Comprehensive National Science Center Hefei230088 China Key Laboratory of Knowledge Engineering with Big Data Hefei University of Technology Ministry of Education Hefei230009 China

Cross-network node classification aims to train a classifier for an unlabeled target network using a source network with rich labels. In applications, the degree of nodes mostly conforms to the long-tail distribution, i.e., most nodes in the network are tail nodes with sparse neighborhoods. The established methods focus on either the discrepancy cross network or the long tail in a single network. As for the cross-network node classification under long tail, the coexistence of sparsity of tail nodes and the discrepancy cross-network challenges existing methods for long tail or methods for the cross-network node classification. To this end, a multicomponent similarity graphs for cross-network node classification (MS-CNC) is proposed in this article. Specifically, in order to address the sparsity of the tail nodes, multiple component similarity graphs, including attribute and structure similarity graphs, are constructed for each network to enrich the neighborhoods of the tail nodes and alleviate the long-tail phenomenon. Then, multiple representations are learned from the multiple similarity graphs separately. Based on the multicomponent representations, a two-level adversarial model is designed to address the distribution difference across networks. One level is used to learn the invariant representations cross network in view of structure and attribute components separately, and the other level is used to learn the invariant representations in view of the fused structure and attribute graphs. Extensive experimental results show that the MS-CNC outperforms the state-of-the-art methods. Impact Statement-Node classification is an important task in graph mining. With the unavailability of labels, some researchers propose cross-network node classification, using one labeled network to assist the node classification of another unlabeled network. However, the long-tail of nodes leads to unsatisfactory performance and challenges the recent cross-network node classification m

关键词： Graphic methods

来源：评论

学校读者我要写书评

暂无评论

Fast Parallel Recovery for Transactional Stream Processing on Multicores 40

Fast Parallel Recovery for Transactional Stream Processing o...

引用

40th IEEE International Conference on data engineering, ICDE 2024

作者： Zhao, Jianjun Liu, Haikun Zhang, Shuhao Duan, Zhuohui Liao, Xiaofei Jin, Hai Zhang, Yu School of Computer Science and Technology Huazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Wuhan China School of Computer Science and Engineering Nanyang Technological University Singapore

ISBN: (纸本)9798350317152

Transactional stream processing engines (TSPEs) have gained increasing attention due to their capability of processing real-time stream applications with transactional semantics. However, TSPEs remain susceptible to system failures and power outages. Existing TSPEs mainly focus on performance improvement, but still face a significant challenge to guarantee fault tolerance while offering high-performance services. We revisit commonly-used fault tolerance approaches in stream processing and database systems, and find that these approaches do not work well on TSPEs due to complex data dependencies. In this paper, we propose a novel TSPE called MorphStreamR to achieve fast failure recovery while guaranteeing low performance overhead at runtime. The key idea of MorphStreamR is to record intermediate results of resolved dependencies at runtime, and thus eliminate data dependencies to improve task parallelism during failure recovery. MorphStreamR further mitigates the runtime overhead by selectively tracking data dependencies and incorporating workload-aware log commitment. Experimental results show that MorphStreamR can significantly reduce the recovery time by up to 3.1 x while experiencing much less performance slowdown at runtime, compared with other applicable fault tolerance approaches. © 2024 IEEE.

关键词： Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

Image Segmentation Survey: Classical and Deep Learning Methods 4

Image Segmentation Survey: Classical and Deep Learning Metho...

引用

4th IEEE International Conference on Electrical, computer, and Energy Technologies, ICECET 2024

作者： Bachani, Viren Roychowdhury, Anish Dasgupta, Abhijit Ei-Alfy, Hazem BSc. Data Science SP Jain School of Global Management Sydney Australia SP Jain School of Global Management Data Science Dept. Sydney Australia SP Jain School of Global Management Data Science Dept. Mumbai India School of Computer Science University of Sydney Sydney Australia Alexandria University Faculty of Engineering Alexandria Egypt

ISBN: (纸本)9798350395914

Image segmentation is a prerequisite to almost all computer vision applications. It enables the extraction of meaningful information from visual inputs by partitioning images into segments with shared features. This review paper begins by outlining different image segmentation tasks. It then surveys a broad spectrum of segmentation techniques, ranging from classical frameworks to deep-learning (DL) architectures. We perform this through a systematic literature analysis across a handful of applications as part of our methodology. A key contribution of our survey is that it elaborates on a few situations where classical approaches outperform DL-based methods in accuracy, along with the added advantage of computational efficiency. In some cases, these findings also provoke interest in revisiting classical methods, prompting strategies that leverage the strengths of both paradigms. © 2024 IEEE.

关键词： Markov processes

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：