检索结果-内蒙古大学图书馆

Dual-View Pyramid Pooling in Deep Neural Networks for Improved Medical Image Classification and Confidence Calibration

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhang, Xiaoqing Nie, Qiushi Xiao, Zunjie Zhao, Jilu Wu, Xiao Guo, Pengxin Li, Runzhi Liu, Jin Wei, Yanjie Pan, Yi Center for High Performance Computing Shenzhen Key Laboratory of Intelligent Bioinformatics Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Guangdong Shenzhen518055 China Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen518055 China Department of Statistics and Actuarial Science The University of Hong Kong 999077 Hong Kong Cooperative Innovation Center of Internet Healthcare Zhengzhou University Zhengzhou450001 China Hunan Provincial Key Lab on Bioinformatics School of Computer Science and Engineering Central South University Changsha410083 China Xinjiang Engineering Research Center of Big Data and Intelligent Software School of software Xinjiang University Wulumuqi830046 China Faculty of Computer Science and Control Engineering Shenzhen University of Advanced Technology Shenzhen China

Spatial pooling (SP) and cross-channel pooling (CCP) operators have been applied to aggregate spatial features and pixel-wise features from feature maps in deep neural networks (DNNs), respectively. Their main goal is to reduce computation and memory overhead without visibly weakening the performance of DNNs. However, SP often faces the problem of losing the subtle feature representations, while CCP has a high possibility of ignoring salient feature representations, which may lead to both miscalibration of confidence issues and suboptimal medical classification results. To address these problems, we propose a novel dual-view framework, the first to systematically investigate the relative roles of SP and CCP by analyzing the difference between spatial features and pixel-wise features. Based on this framework, we propose a new pooling method, termed dual-view pyramid pooling (DVPP), to aggregate multi-scale dual-view features. DVPP aims to boost both medical image classification and confidence calibration performance by fully leveraging the merits of SP and CCP operators from a dual-axis perspective. Additionally, we discuss how to fulfill DVPP with five parameter-free implementations. Extensive experiments on six 2D/3D medical image classification tasks show that our DVPP surpasses state-of-the-art pooling methods in terms of medical image classification results and confidence calibration across different DNNs. © 2024, CC BY.

关键词： Deep neural networks

SPA: A Graph Spectral Alignment Perspective for Domain Adaptation

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Xiao, Zhiqing Wang, Haobo Jin, Ying Feng, Lei Chen, Gang Huang, Fei Zhao, Junbo College of Computer Science and Technology Zhejiang University China School of Software Technology Zhejiang University China Key Lab of Intelligent Computing based Big Data of Zhejiang Province Zhejiang University China CUHK-SenseTime Joint Lab The Chinese University of Hong Kong Hong Kong School of Computer Science and Engineering Nanyang Technological University Singapore Alibaba Group China

Unsupervised domain adaptation (UDA) is a pivotal form in machine learning to extend the in-domain model to the distinctive target domains where the data distributions differ. Most prior works focus on capturing the inter-domain transferability but largely overlook rich intra-domain structures, which empirically results in even worse discriminability. In this work, we introduce a novel graph SPectral Alignment (SPA) framework to tackle the tradeoff. The core of our method is briefly condensed as follows: (i)-by casting the DA problem to graph primitives, SPA composes a coarse graph alignment mechanism with a novel spectral regularizer towards aligning the domain graphs in eigenspaces;(ii)-we further develop a fine-grained message propagation module — upon a novel neighbor-aware self-training mechanism — in order for enhanced discriminability in the target domain. On standardized benchmarks, the extensive experiments of SPA demonstrate that its performance has surpassed the existing cutting-edge DA methods. Coupled with dense model analysis, we conclude that our approach indeed possesses superior efficacy, robustness, discriminability, and transferability. Code and data are availab.e at: https://***/CrownX/SPA. Copyright © 2023, The Authors. All rights reserved.

关键词： Benchmarking

Federated Linear Bandit Learning via Over-the-air Computation

学校读者我要写书评

暂无评论

Federated Linear Bandit Learning via Over-the-air Computatio...

IEEE Conference on Global Communications (GLOBECOM)

作者： Jiali Wang Yuning Jiang Xin Liu Ting Wang Yuanming Shi MoE Engineering Research Center of Software/Hardware Co-design Technology and Application Shanghai Key Lab. of Trustworthy Computing East China Normal University China School of Information Science and Technology (SIST) ShanghaiTech University China Automatic Control Laboratory Ecole Polytechnique Fédérale de Lausanne (EPFL) Switzerland

In this paper, we investigate federated contextual linear bandit learning within a wireless system that comprises a server and multiple devices. Each device interacts with the environment, selects an action based on the received reward, and sends model updates to the server. The primary objective is to minimize cumulative regret across all devices within a finite time horizon. To reduce the communication overhead, devices communicate with the server via over-the-air computation (AirComp) over noisy fading channels, where the channel noise may distort the signals. In this context, we propose a customized federated linear bandits scheme, where each device transmits an analog signal, and the server receives a superposition of these signals distorted by channel noise. A rigorous mathematical analysis is conducted to determine the regret bound of the proposed scheme. Both theoretical analysis and numerical experiments demonstrate the competitive performance of our proposed scheme in terms of regret bounds in various settings.

关键词：

Federated Reinforcement Learning for Electric Vehicles Charging Control on Distribution Networks

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Qian, Junkai Jiang, Yuning Liu, Xin Wang, Qiong Wang, Ting Shi, Yuanming Chen, Wei The MoE Engineering Research Center of Software/Hardware Co-design Technology and Application The Shanghai Key Lab. of Trustworthy Computing East China Normal University China The Automatic Control Laboratory EPFL Switzerland The School of Information Science and Technology ShanghaiTech University China The State Grid Beijing Electric Power Company China The Department of Electronic Engineering Tsinghua University China

With the growing popularity of electric vehicles (EVs), maintaining power grid stability has become a significant challenge. To address this issue, EV charging control strategies have been developed to manage the switch between vehicle-to-grid (V2G) and grid-to-vehicle (G2V) modes for EVs. In this context, multi-agent deep reinforcement learning (MADRL) has proven its effectiveness in EV charging control. However, existing MADRL-based approaches fail to consider the natural power flow of EV charging/discharging in the distribution network and ignore driver privacy. To deal with these problems, this paper proposes a novel approach that combines multi-EV charging/discharging with a radial distribution network (RDN) operating under optimal power flow (OPF) to distribute power flow in real time. A mathematical model is developed to describe the RDN load. The EV charging control problem is formulated as a Markov Decision Process (MDP) to find an optimal charging control strategy that balances V2G profits, RDN load, and driver anxiety. To effectively learn the optimal EV charging control strategy, a federated deep reinforcement learning algorithm named FedSAC is further proposed. Comprehensive simulation results demonstrate the effectiveness and superiority of our proposed algorithm in terms of the diversity of the charging control strategy, the power fluctuations on RDN, the convergence efficiency, and the generalization ability. © 2023, CC BY-NC-SA.

关键词： Reinforcement learning

AIDE: Attack Inference Based on Heterogeneous Dependency Graphs with MITRE ATT&CK

学校读者我要写书评

暂无评论

AIDE: Attack Inference Based on Heterogeneous Dependency Gra...

IEEE International Conference on Trust, Security and Privacy in computing and Communications (TrustCom)

作者： Weidong Zhou Chunhe Xia Nan Feng Xinyi Pan Tianbo Wang Xiaojian Li Beijing Key Lab. of Network Technology Beihang University Beijing China Guangxi Collaborative Innovation Center of Multi-Source Information Integration and Intelligent Processing Guangxi Normal University Guilin China School of Software & Microelectronics Peking University Beijing China School of Cyber Science and Technology Beihang University Beijing China School of Computer Science and Engineering Guangxi Normal University Guilin China

ISBN: (数字)9798331506209

ISBN: (纸本)9798331506216

The MITRE ATT&CK Framework provides a rich and actionable repository of adversarial tactics, techniques, and procedures (TTPs). By leveraging the associations among these techniques, we can infer unobserved techniques based on those have been observed. Current machine learning inference methods based on small-scale data face the following problems: "logical dependencies missing" and "complex associations missing." The former means that current association methods rely solely on the statistical dependencies between attack techniques while ignoring the logical dependencies involved in executing those techniques. The latter indicates that current inference methods simplify the complex associations between technologies based on statistical assumptions, failing to adequately capture the long-range and intricate real-world relationships between techniques. To address these problems, we propose an Attack Inference model based on heterogeneous DEpendency graphs (AIDE). First, we extract three types of logical dependencies: CVE, Pattern, and Conditional, which allows us to construct a heterogeneous dependency graph among the techniques, thus addressing the "logical dependencies missing" problem. Second, we employ Relation Graph Convolution Neural Networks (RGCN) to aggregate heterogeneous dependencies into representation vectors of the techniques. Finally, we utilize the sequence inference model Decoder, which excels at capturing implicit contextual associations, to infer unknown attacks, thus addressing the "complex associations missing" problem. Experiments conducted on real datasets demonstrate that our model significantly outperforms state-of-the-art baseline models in both hits@k and map@k metrics.

关键词： Adaptation models Aggregates Refining Machine learning Vectors Threat assessment Real-time systems Decoding Security Context modeling

DivGCL: A Graph Contrastive Learning Model for Diverse Recommendation 39

学校读者我要写书评

暂无评论

DivGCL: A Graph Contrastive Learning Model for Diverse Recom...

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Gong, Wenwen Geng, Yangliao Zhang, Dan Zhu, Yifan Xu, Xiaolong Xiang, Haolong Beheshti, Amin Zhang, Xuyun Qi, Lianyong School of Information and Electrical Engineering China Agricultural University China School of Computer Science & Technology Beijing Jiaotong University China Department of Computer Science and Technology Tsinghua University China School of Computer Science Beijing University of Posts and Telecommunications China School of Software Nanjing University of Information Science and Technology China School of Computing Macquarie University Australia College of Computer Science and Technology China University of Petroleum (East China China State Key Lab. for Novel Software Technology Nanjing University China

ISBN: (纸本)157735897X

Graph Contrastive Learning (GCL), as a primary paradigm of graph self-supervised learning, spurs a fruitful line of research in tackling the data sparsity issue by maximizing the consistency of user/item embeddings between different augmented views with random perturbations. However, diversity, as a crucial metric for recommendation performance and user satisfaction, has received rather little attention. In fact, there exists a challenging dilemma in balancing accuracy and diversity. To address these issues, we propose a new Graph Contrastive Learning model for Diversifying recommendations (DivGCL). Inspired by the excellence of the determinant point process (DPP), DivGCL adopts a DPP likelihood-based loss function to achieve an ideal trade-off between diversity and accuracy, optimizing it jointly with the advanced Gaussian noise-augmented GCL objective. Extensive experiments on four popular datasets demonstrate that DivGCL surpasses existing approaches in balancing accuracy and diversity, with an improvement of 23.47% at T@20 (abbreviation for trade-off metric) on ML-1M. Copyright © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Self-supervised learning

Over-the-Air Federated Learning via Second-Order Optimization

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yang, Peng Jiang, Yuning Wang, Ting Zhou, Yong Shi, Yuanming Jones, Colin N. The Shanghai Key Lab. of Trustworthy Computing Software Engineering Institute East China Normal University Shanghai200062 China The Automatic Control Laboratory EPFL Lausanne1015 Switzerland The School of Information Science and Technology ShanghaiTech University Shanghai201210 China

Federated learning (FL) is a promising learning paradigm that can tackle the increasingly prominent isolated data islands problem while keeping users' data locally with privacy and security guarantees. However, FL could result in task-oriented data traffic flows over wireless networks with limited radio resources. To design communication-efficient FL, most of the existing studies employ the first-order federated optimization approach that has a slow convergence rate. This however results in excessive communication rounds for local model updates between the edge devices and edge server. To address this issue, in this paper, we instead propose a novel over-the-air second-order federated optimization algorithm to simultaneously reduce the communication rounds and enable low-latency global model aggregation. This is achieved by exploiting the waveform superposition property of a multi-access channel to implement the distributed second-order optimization algorithm over wireless networks. The convergence behavior of the proposed algorithm is further characterized, which reveals a linear-quadratic convergence rate with an accumulative error term in each iteration. We thus propose a system optimization approach to minimize the accumulated error gap by joint device selection and beamforming design. Numerical results demonstrate the system and communication efficiency compared with the state-of-the-art approaches. Copyright © 2022, The Authors. All rights reserved.

关键词： Wireless networks

Harnessing Hierarchical lab.l Distribution Variations in Test Agnostic Long-tail Recognition

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yang, Zhiyong Xu, Qianqian Wang, Zitai Li, Sicong Han, Boyu Bao, Shilong Cao, Xiaochun Huang, Qingming School of Computer Science and Tech. University of Chinese Academy of Sciences China Key Lab. of Intelligent Information Processing Institute of Computing Tech. CAS Institute of Information Engineering CAS School of Cyber Security University of Chinese Academy of Sciences China School of Cyber Science and Tech. Shenzhen Campus of Sun Yat-sen University China BDKM University of Chinese Academy of Sciences China

This paper explores test-agnostic long-tail recognition, a challenging long-tail task where the test lab.l distributions are unknown and arbitrarily imbalanced. We argue that the variation in these distributions can be broken down hierarchically into global and local levels. The global ones reflect a broad range of diversity, while the local ones typically arise from milder changes, often focused on a particular neighbor. Traditional methods predominantly use a Mixture-of-Expert (MoE) approach, targeting a few fixed test lab.l distributions that exhibit substantial global variations. However, the local variations are left unconsidered. To address this issue, we propose a new MoE strategy, DirMixE, which assigns experts to different Dirichlet meta-distributions of the lab.l distribution, each targeting a specific aspect of local variations. Additionally, the diversity among these Dirichlet meta-distributions inherently captures global variations. This dual-level approach also leads to a more stable objective function, allowing us to sample different test distributions better to quantify the mean and variance of performance outcomes. Theoretically, we show that our proposed objective benefits from enhanced generalization by virtue of the variance-based regularization. Comprehensive experiments across multiple benchmarks confirm the effectiveness of DirMixE. The code is availab.e at https: //***/scongl/DirMixE. © 2024, CC BY.

关键词：

A Novel Offloading Approach of Iot User Perception Task Based on Quantum Behavior Particle Swarm Optimization

学校读者我要写书评

暂无评论

SSRN

SSRN 2022年

作者： Chen, Lu Zhang, De-Gan Zhang, Jie Zhang, Ting Wang, Wen-Jing Cao, Ya-Hui Tianjin Key Lab of Intelligent Computing and Novel Software Technology Tianjin University of Technology Tianjin300384 China School of Electronic and Information Engineering Beijing Jiaotong University Beijing100044 China School of Sports Economics and Management Tianjin University of Sport Tianjin301617 China

To solve the problem of high time consumption and energy loss of edge computing perceptual task offloading in the mobile IoT (Internet of things), it is necessary to increase the particle population diversity and the performance of escaping the local optimum solution in the intelligent optimization algorithm. For the offloading scenario of multi edge central servers and smart devices of multi IoT users, the NQPSOOM method (A new method of user perceptual task offloading in the IoT based on optimizing strategy of quantum behavior particle population) was proposed. The adaptability variance of the quantum behavior particle population was designed to measure the convergence of the particle population. The logistic chaos perturbation strategy was used to improve the performance of the intelligent optimization algorithm of escaping the local optimum solution. The exchange and mutation operations of fast elite non-dominated sorting genetic algorithm based on crowding coefficient were designed to select the particle population with a high diversity of quantum behavior particle population. Compared with the other five perceptual task computing offloading methods in recent three years, the NQPSOOM method reduces the adaptability of the intelligent optimization algorithm, decreases the time consumption and energy loss of the algorithm. © 2022, The Authors. All rights reserved.

关键词： Internet of things