检索结果-内蒙古大学图书馆

distributed robust support vector ordinal regression under label noise

NEUROCOMPUTING 2024年 598卷

作者： Liu, Huan Tu, Jiankai Gao, Anqi Li, Chunguang Zhejiang Univ Coll Informat Sci & Elect Engn Hangzhou 310027 Peoples R China Univ Nottingham Ningbo China Sch Comp Sci Ningbo 315199 Peoples R China

Ordinal regression (OR) methods are designed for a type of classification problems where data labels have natural orders. In practice, data may be corrupted by label noise, which affects the training process thus degrading the generalization performance of OR methods. In OR, data are usually assumed to have latent variables underlying the ordinal labels, and label noise exhibits a special characteristic that it usually causes large latent variable value deviations. However, there are few existing works on OR considering label noise, and the existing works do not utilize the above -mentioned characteristic. Besides, most of the existing OR methods are centralized, which are inapplicable in some realistic distributed applications. In this paper, we utilize the characteristic of label noise in OR to develop a distributed robust support vector ordinal regression method (dRSVOR) under label noise. Specifically, after analyzing the characteristic of label noise in OR, we take the form of SVOR with explicit constraints to achieve robustness to one type of mislabeled samples. Then, we adopt correntropy, an information -theoretic measure, to achieve robustness to the other type of mislabeled samples. Theoretically, we analyze the consensus and convergence of dRSVOR. Experimentally, we conduct experiments on both synthetic data and real OR datasets to illustrate the effectiveness of the proposed method. The results show that the centralized version of dRSVOR outperforms several state-of-the-art OR methods considering label noise in centralized circumstances with label noise, and dRSVOR could approach the performance of the centralized version despite additional constraints in distributed scenarios.

关键词： Ordinal regression Support vector ordinal regression Label noise Correntropy distributed algorithm

来源：评论

学校读者我要写书评

暂无评论

Cooperative coevolution for non-separable large-scale black-box optimization: Convergence analyses and distributed accelerations

引用

APPLIED SOFT COMPUTING 2024年 166卷

作者： Duan, Qiqi Shao, Chang Zhou, Guochen Yang, Haobin Zhao, Qi Shi, Yuhui Harbin Inst Technol Harbin 150001 Heilongjiang Peoples R China Univ Technol Sydney Australian Artificial Intelligence Inst AAII Fac Engn & Informat Technol 15 Broadway Sydney NSW 2007 Australia Southern Univ Sci & Technol Dept Comp Sci & Engn Shenzhen 518055 Guangdong Peoples R China

Given the ubiquity of non-separable optimization problems in real worlds, in this paper we analyze and extend the large-scale version of the well-known cooperative coevolution (CC), a divide-and-conquer black-box optimization framework, on non-separable functions. First, we reveal empirical reasons of when decomposition- based methods are preferred or not in practice on some non-separable large-scale problems, which have not been clearly pointed out in many previous CC papers. Then, we formalize CC to a continuous-game model via simplification, but without losing its essential property. Different from previous evolutionary game theory for CC, our new model provides a much simpler but useful viewpoint to analyze its convergence, since only the pure Nash equilibrium concept is needed and more general fitness landscapes can be explicitly considered. Based on convergence analyses, we propose a hierarchical decomposition strategy for better generalization, as for any decomposition, there is a risk of getting trapped into a suboptimal Nash equilibrium. Finally, we use powerful distributed computing to accelerate it under the recent multi-level learning framework, which combines the fine-tuning ability from decomposition with the invariance property of CMA-ES. Experiments on a set of high-dimensional test functions validate both its search performance and scalability (w.r.t. CPU cores) on a clustering computing platform with 400 CPU cores.

关键词： Black-box optimization Convergence analysis Cooperative coevolution distributed algorithm Large-scale optimization

来源：评论

学校读者我要写书评

暂无评论

distributed Nash Equilibrium Seeking for Aggregative Games over Weight-Unbalanced Digraphs 16th

Distributed Nash Equilibrium Seeking for Aggregative Games o...

引用

16th IFAC Symposium on Large Scale Complex Systems - Theory and Applications (LSS)

作者： Wang, Dong Chen, Pan Chen, Mingfei Lian, Jie Wang, Wei Dalian Univ Technol Sch Control Sci & Engn Minist Educ Key Lab Intelligent Control & Optimizat Ind Equip Dalian Peoples R China

This paper investigates a Nash equilibrium seeking problem of aggregative games with linear coupling constraints, where the cost function of each player is coupled through an aggregative function. All players seek the Nash equilibrium by exchanging decision information with neighbors over strongly connected and weight-unbalanced digraphs. A novel distributed continuous-time algorithm is designed by using the Lagrangian multiplier method and the dynamic average consensus method. Then, we show exponential convergence of the designed algorithm by utilizing stability theory of nonlinear systems. Finally, a numerical simulation is provided to verify the effectiveness of our algorithm. Copyright (c) 2022 The Authors. This is an open access article under the CC BY-NC-ND License (http://***/licenses/by-nc-nd/4.0/)

关键词： distributed algorithm aggregative games coupling constraints digraphs

来源：评论

学校读者我要写书评

暂无评论

Multi-commodity Optimization of Peer-to-peer Energy Trading Resources in Smart Grid

引用

Journal of Modern Power Systems and Clean Energy 2022年第1期10卷 29-39页

作者： Olamide Jogunola Bamidele Adebisi Kelvin Anoh Augustine Ikpehai Mohammad Hammoudeh Georgina Harris Manchester Metropolitan University ManchesterUK Department of Electrical and Electronic Engineering University of BoltonBoltonUK Department of Engineering and Mathematics Sheffield Hallam UniversitySheffieldUK Department of Computing and Mathematics Manchester Metropolitan UniversityManchesterUK School of Science Engineering and EnvironmentUniversity of SalfordSalfordUK

Utility maximization is a major priority of prosumers participating in peer-to-peer energy trading and sharing(P2P-ETS).However,as more distributed energy resources integrate into the distribution network,the impact of the communication link becomes *** present a multi-commodity formulation that allows the dual-optimization of energy and communication resources in *** one hand,the proposed algorithm minimizes the cost of energy generation and communication *** the other hand,it also maximizes the global utility of prosumers with fair resource *** evaluate the algorithm in a variety of realistic conditions including a time-varying communication network with signal delay signal *** results show that the convergence is achieved in a fewer number of time steps than the previously proposed *** is further observed that the entities with a higher willingness to trade the energy acquire more satisfactions than others.

关键词： distributed algorithm social welfare peer-to-peer energy trading and sharing multi-commodity networks economic dispatch packet loss peer-to-peer energy trading distributed dual-gradient(DDG)

来源：评论

学校读者我要写书评

暂无评论

Optimizing Spectrum Usage in Dynamic Wireless Mesh Networks: A Cognitive Radio Approach 4

Optimizing Spectrum Usage in Dynamic Wireless Mesh Networks:...

引用

4th IEEE Interdisciplinary Conference on Electrics and Computer (INTCEC)

作者： Tagare, Trupti Shripad Vidyashree, K. N. Chaitra, A. Kumar, Mahesh N. Kumar, Drava S. Shruthi, M. Dayananda Sagar Coll Engn Dept ECE Bengaluru India

ISBN: (纸本)9798350349467;9798350349450

Adaptable topologies and dynamic node arrangements are essential for sustaining connectivity in wireless mesh networks and guaranteeing continuous internet access. Solving basic networking problems, such as the shortest path problem, requires the use of multiple techniques and algorithms. To improve packet transmission by utilizing optimal node selections to reduce loss and latency, this study uses Dijkstra's approach to determine the most effective paths inside the network. Moreover, the main problem with current wireless networks is the use of radio resources. The development of cognitive radio networks has emerged as a potential remedy to improve spectrum use. Allocating spectrum efficiently is essential to improving the performance of cognitive networks. This paper explores the potential integration of cognitive radio principles into wireless mesh networks, providing insights into spectrum optimization techniques to improve network dependability and efficiency.

关键词： cognitive mesh network distributed algorithm gaming theory channel spectrum

来源：评论

学校读者我要写书评

暂无评论

distributed real-time pricing of smart grid considering individual differences

引用

OMEGA-INTERNATIONAL JOURNAL OF MANAGEMENT SCIENCE 2024年 127卷

作者： Qu, Deqiang Li, Junxiang Ma, Xiaojia Univ Shanghai Sci & Technol Business Sch Shanghai 200093 Peoples R China Univ Shanghai Sci & Technol Sch Intelligent Emergency Management Shanghai 200093 Peoples R China

The utility function that characterizes customers' satisfaction with electricity consumption plays an important and irreplaceable role in the real-time pricing mechanism based on the social welfare maximization model. Without the accurate quantification of customers' utility, the real-time pricing will deviate from the reality. In fact, the utility functions of different types of customers in different regions should be obtained by fitting a large amount of historical data over a long period of time based on insight into the relationship between factors and utility. Based on the consideration of customers' individual differences, a new utility function is proposed, which enriches the form of the utility function and provides a reference for fitting a real and accurate utility function. Further, based on this proposed utility function, a real-time pricing model of social welfare maximization is developed to obtain the fair electricity price between customers and the power supplier. On the basis of the separable structure of variables, we design distributed algorithms with global convergence for the pricing model and estimate a worst -case convergence rate. Numerical simulations verify the feasibility and effectiveness of our algorithms and the rationality of the new utility function, i.e., the electricity price based on the proposed utility function is more robust than the existing ones.

关键词： Real-time pricing Social welfare maximization distributed algorithm Robustness

来源：评论

学校读者我要写书评

暂无评论

distributed Fixed-Time Optimization in Economic Dispatch Over Directed Networks

引用

IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS 2021年第5期17卷 3011-3019页

作者： Dai, Hao Jia, Jinping Yan, Li Fang, Xinpeng Chen, Weisheng Xidian Univ Sch Aerosp Sci & Technol Xian 710071 Peoples R China Tianshui Normal Univ Sch Math & Stat Tianshui 741001 Peoples R China Xidian Univ Xian 710071 Peoples R China

A distributed algorithm is presented in this article under directed communication networks, which is used to solve the economic dispatch problem in fixed time in smart grid systems. A new globally fixed-time stability theory (Lemma 3) is first given in this article, which contains a new upper bound for the estimation of the settling time. Moreover, the fixed-time convergence for the proposed algorithm is rigorously proved with the aid of convex optimization theory, the new lemma, and Lyapunov stability theory under the strongly connected and weight-balanced network topology. Finally, numerical simulations show the effectiveness and advantages of the distributed fixed-time optimization algorithm.

关键词： Optimization Economics Smart grids Convex functions Informatics Communication networks distributed algorithms Convex optimization directed network distributed algorithm equality constraint fixed-time convergence

来源：评论

学校读者我要写书评

暂无评论

distributed Decision Making for V2V Charge Sharing in Intelligent Transportation Systems 19

Distributed Decision Making for V2V Charge Sharing in Intell...

引用

19th Annual IEEE International Conference on Sensing, Communication, and Networking (IEEE SECON)

作者： Chatterjee, Punyasha Majumder, Pratham Debnath, Arpita Das, Sajal K. Jadavpur Univ Sch Mobile Comp Commun Kolkata India CMR Inst Technol Dept Comp Sci & Engn Bengaluru India Calcutta Tech Sch Dept Comp Sci & Technol Kolkata India Missouri Univ Sci & Technol Dept Comp Sci Rolla MO 65409 USA

ISBN: (数字)9781665486439

ISBN: (纸本)9781665486439

Electric vehicles (EVs) have emerged in the intelligent transportation system (ITS) to meet the increasing environmental concerns. To facilitate on-demand requirement of EV charging, vehicle-to-vehicle (V2V) charge transfer can be employed. However, most of the existing approaches to V2V charge sharing are centralized or semi-centralized, incurring huge message overhead, long waiting time, and infrastructural cost. In this paper, we propose novel distributed heuristic algorithms for V2V charge sharing based on the multi-criteria decision-making policy. The problem is mapped to an alias classical problem (i.e., optimum matching in weighted bipartite graphs), where the goal is to maximize the matching cardinality while minimizing the matching cost. An integer linear programming (ILP)-based problem formulation can not achieve optimum matching because the global network topology is not available with the EVs due to their limited communication range. Our proposed heuristics can yield an almost stable matching with lesser computational and message overhead compared to other existing distributed approaches. An average case matching probability is also calculated. Simulation experiments are conducted to measure the performance of our heuristics in terms of message overhead, matching percentage, and matching preference. The proposed solution outperforms the existing distributed approaches and shows comparable result with respect to standard centralized stable matching algorithm.

关键词： Intelligent Transportation System (ITS) Electric Vehicles (EVs) V2V charge sharing multi-criteria decision making distributed algorithm matching

来源：评论

学校读者我要写书评

暂无评论

Fault-tolerant Local Recovery with Preprocessing in Multiple Shared Protection 14

Fault-tolerant Local Recovery with Preprocessing in Multiple...

引用

14th International Workshop on Resilient Networks Design and Modeling

作者： Zhu, Mengfei Kang, Rui Foerster, Klaus-Tycho China Mobile Grp Design Inst Co Ltd Beijing Peoples R China Kyoto Univ Kyoto Japan TU Dortmund Dortmund Germany

ISBN: (纸本)9798331518189;9798331518196

In distributed microservice architecture, low latency and high reliability are crucial as they directly impact user experience by ensuring robust service continuity across the network. Function unavailability occurs due to various reasons, and when such failures are detected, the nodes hosting the backup resources can attempt autonomously local recovery, leveraging precomputed recovery policies, without the need for central processor intervention, thereby reducing recovery time. In scenarios involving shared protection with limited recovery capabilities, it is crucial to implement a well-defined policy for local recovery. Such a policy prevents repetitive or infeasible recovery attempts, ensuring effective recovery. This paper considers a local strategy to maximize the residual resilience, where individual nodes decide which function to recover. It hence works as a complement to global recovery to provide prompt recovery, as a first line of defense. This paper hence considers the preparation of priority settings with preprocessing based on the SUPPORTED *** conduct a comprehensive analysis of the conditions for achieving feasible and optimal local recovery solutions with zero-round communication for one and two failures. Additionally, we formulate the problem as an integer linear programming problem and design a testbed to implement the local recovery algorithm incorporating preprocessing. Our evaluation reveals that the zero-round communication approach emerges as the optimal choice, drastically shortening recovery times to 10 seconds while maintaining 98% feasibility and perfect optimality for all practical solutions, on average, across two failure scenarios.

关键词： Local recovery preprocessing distributed algorithm shared protection fault tolerance residual resilience network virtualization

来源：评论

学校读者我要写书评

暂无评论

distributed algorithms for Fully Personalized PageRank on Large Graphs 19

Distributed Algorithms for Fully Personalized PageRank on La...

引用

World Wide Web Conference (WWW)

作者： Lin, Wenqing Tencent Inc Interact Entertainment Grp Shenzhen Peoples R China

ISBN: (纸本)9781450366748

Personalized PageRank (PPR) has enormous applications, such as link prediction and recommendation systems for social networks, which often require the fully PPR to be known. Besides, most of real-life graphs are edge-weighted, e.g., the interaction between users on the Facebook network. However, it is computationally difficult to compute the fully PPR, especially on large graphs, not to mention that most existing approaches do not consider the weights of edges. In particular, the existing approach cannot handle graphs with billion edges on a moderate-size cluster. To address this problem, this paper presents a novel study on the computation of fully edge-weighted PPR on large graphs using the distributed computing framework. Specifically, we employ the Monte Carlo approximation that performs a large number of random walks from each node of the graph, and exploits the parallel pipeline framework to reduce the overall running time of the fully PPR. Based on that, we develop several optimization techniques which (i) alleviate the issue of large nodes that could explode the memory space, (ii) pre-compute short walks for small nodes that largely speedup the computation of random walks, and (iii) optimize the amount of random walks to compute in each pipeline that significantly reduces the overhead. With extensive experiments on a variety of real-life graph datasets, we demonstrate that our solution is several orders of magnitude faster than the state-of-the-arts, and meanwhile, largely outperforms the baseline algorithms in terms of accuracy.

关键词： distributed algorithm Graph Personalized PageRank

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：