检索结果-内蒙古大学图书馆

Method to simultaneously facilitate all jet physics tasks

学校读者我要写书评

暂无评论

Physical Review D 2025年第5期111卷 054015-054015页

作者： Vinicius Mikuni Benjamin Nachman National Energy Research Scientific Computing Center Berkeley Lab Berkeley California 94720 USA Physics Division Lawrence Berkeley National Laboratory Berkeley California 94720 USA Berkeley Institute for Data Science University of California Berkeley California 94720 USA

Machine learning has become an essential tool in jet physics. Due to their complex, high-dimensional nature, jets can be explored holistically by neural networks in ways that are not possible manually. However, innovations in all areas of jet physics are proceeding in parallel. We show that specially constructed machine learning models trained for a specific jet classification task can improve the accuracy, precision, or speed of all other jet physics tasks. This is demonstrated by training on a particular multiclass generation and classification task and then using the learned representation for different generation and classification tasks, for datasets with a different (full) detector simulation, for jets from a different collision system (pp versus ep), for generative models, for likelihood ratio estimation, and for anomaly detection. We consider our omnilearn approach thus as a jet-physics foundation model. It is made publicly available for use in any area where state-of-the-art precision is required for analyses involving jets and their substructure.

关键词：

ACE-RLHF: Automated Code Evaluation and Socratic Feedback Generation Tool using Large Language Models and Reinforcement Learning with Human Feedback

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Rahman, Tasnia Kumar, Sathish A.P. Jha, Sumit Ramanathan, Arvind Department of Computer Science Cleveland State University ClevelandOH United States School of Computing and Information Sciences Florida International University MiamiFL United States Data Science and Learning Division Argonne National Laboratory LemontIL United States

Automated Program Repair tools are developed for generating feedback and suggesting a repair method for erroneous code. State of the art (SOTA) code repair methods rely on data-driven approaches and often fail to deliver solution for complicated programming questions. To interpret the natural language of unprecedented programming problems, using Large Language Models (LLMs) for code-feedback generation is crucial. LLMs generate more comprehensible feedback than compiler-generated error messages, and Reinforcement Learning with Human Feedback (RLHF) further enhances quality by integrating human-in-the-loop which helps novice students to lean programming from scratch interactively. We are applying RLHF fine-tuning technique for an expected Socratic response such as a question with hint to solve the programming issue. We are proposing code feedback generation tool by fine-tuning LLM with RLHF, Automated Code Evaluation with RLHF (ACE-RLHF), combining two open-source LLM models with two different SOTA optimization techniques. The quality of feedback is evaluated on two benchmark datasets containing basic and competition-level programming questions where the later is proposed by us. We achieved 2-5% higher accuracy than RL-free SOTA techniques using Llama-3-7B-Proximal-policy optimization in automated evaluation and similar or slightly higher accuracy compared to reward model-free RL with AI Feedback (RLAIF). We achieved almost 40% higher accuracy with GPT-3.5 Best-of-n optimization while performing manual evaluation. © 2025, CC BY.

关键词： Reinforcement learning

Enhanced Penalty-based Bidirectional Reinforcement Learning Algorithms

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Pula, Gana Sandeep Kumar, Sathish A.P. Jha, Sumit Ramanathan, Arvind Department of Computer Science Cleveland State University ClevelandOH United States School of Computing and Information Sciences Florida International University MiamiFL United States Data Science and Learning Division Argonne National Laboratory LemontIL United States

This research focuses on enhancing reinforcement learning (RL) algorithms by integrating penalty functions to guide agents in avoiding unwanted actions while optimizing rewards. The goal is to improve the learning process by ensuring that agents learn not only suitable actions but also which actions to avoid. Additionally, we reintroduce a bidirectional learning approach that enables agents to learn from both initial and terminal states, thereby improving speed and robustness in complex environments. Our proposed Penalty-Based Bidirectional methodology is tested against Mani skill benchmark environments, demonstrating an optimality improvement of success rate of approximately 4% compared to existing RL implementations. The findings indicate that this integrated strategy enhances policy learning, adaptability, and overall performance in challenging scenarios. © 2025, CC BY.

关键词： Learning algorithms

Dynamic Charging and Path Planning for UAV-Powered Rechargeable WSNs Using Multi-Agent Deep Reinforcement Learning

学校读者我要写书评

暂无评论

IEEE Transactions on Automation science and Engineering 2025年 22卷 15610-15626页

作者： Betalo, Mesfin Leranso Leng, Supeng Seid, Abegaz Mohammed Abishu, Hayla Nahom Erbad, Aiman Bai, Xiaoshan Shenzhen University College of Mechatronics and Control Engineering Shenzhen518060 China University of Electronic Science and Technology of China School of Information and Communication Engineering Chengdu611731 China Hamad Bin Khalifa University Division of Information and Computing Technology College of Science and Engineering Doha Qatar Shenzhen University College of Mechatronics and Control Engineering National Engineering Laboratory for Big Data System Computing Technology Shenzhen518060 China

Unmanned Aerial Vehicle (UAV)-powered 5G/6G networks integrated with rechargeable wireless sensor networks (RWSNs) offer promising solutions for extending system lifetime, collecting data, and providing computing services and power to sensor nodes (SNs). UAVs offer significant advantages, including exceptional mobility, cost-effective deployment, and the ability to be easily reprogrammed for a wide range of missions. However, the limited onboard power capacity of UAVs, coupled with the lack of dynamic and intelligent charging station (CS) management and inefficient path planning, can lead to SN failure in dynamic mobile environments. To address these challenges, we propose an energy-efficient laser-charged UAV (LCU)-enabled RWSN environment, wherein UAVs, powered by laser beams from ground-based stations, provide services, collect data, and transfer energy to SNs. We formulate a joint optimization problem involving power allocation, dynamic charging strategy (DCS), and path planning to minimize task completion time and sensor node death time. Given the NP-hard nature of the problem, we employ a stochastic game model based on a Markov decision process (MDP) for its solution. To solve this problem, we propose a deep reinforcement learning (DRL) based algorithm that enables real-time charging scheduling decisions while optimizing network performance. We introduce a multi-agent double deep Q-network (MA-DDQN) model to determine the optimal trajectories for all UAVs in large and complex environments. Simulation results demonstrate that the MA-DDQN approach outperforms state-of-the-art techniques, showing significant improvements in terms of average delay, energy consumption, and task completion time. © 2004-2012 IEEE.

关键词： Stochastic models

FedMUA: Exploring the Vulnerabilities of Federated Learning to Malicious Unlearning Attacks

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Chen, Jian Lin, Zehui Lin, Wanyu Shi, Wenlong Yin, Xiaoyan Wang, Di Department of Data Science and Artificial Intelligence The Hong Kong Polytechnic University Hong Kong Department of Data Science and Artificial Intelligence The Department of Computing The Hong Kong Polytechnic University Hong Kong School of Electronic Information and Communications Huazhong University of Science and Technology Wuhan430074 China School of Information Science and Technology Northwest University Xi’an China Division of Computer Electrical and Mathematical Sciences and Engineering King Abdullah University of Science and Technology Saudi Arabia

Recently, the practical needs of "the right to be forgotten" in federated learning gave birth to a paradigm known as federated unlearning, which enables the server to forget personal data upon the client’s removal request. Existing studies on federated unlearning have primarily focused on efficiently eliminating the influence of requested data from the client’s model without retraining from scratch, however, they have rarely doubted the reliability of the global model posed by the discrepancy between its prediction performance before and after unlearning. To bridge this gap, we take the first step by introducing a novel malicious unlearning attack dubbed FedMUA, aiming to unveil potential vulnerabilities emerging from federated learning during the unlearning process. Specifically, clients may act as attackers by crafting malicious unlearning requests to manipulate the prediction behavior of the global model. The crux of FedMUA is to mislead the global model into unlearning more information associated with the influential samples for the target sample than anticipated, thus inducing adverse effects on target samples from other clients. To achieve this, we design a novel two-step method, known as Influential Sample Identification and Malicious Unlearning Generation, to identify and subsequently generate malicious feature unlearning requests within the influential samples. By doing so, we can significantly alter the predictions pertaining to the target sample by initiating the malicious feature unlearning requests, leading to the deliberate manipulation for the user adversely. Additionally, we design a new defense mechanism that is highly resilient against malicious unlearning attacks. Extensive experiments on three realistic datasets reveal that FedMUA effectively induces misclassification on target samples and can achieve an 80% attack success rate by triggering only 0.3% malicious unlearning requests. Copyright © 2025, The Authors. All rights reserved.

关键词： Federated learning

SACB-Net: Spatial-awareness Convolutions for Medical Image Registration

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Cheng, Xinxing Zhang, Tianyang Lu, Wenqi Meng, Qingjie Frangi, Alejandro F. Duan, Jinming School of Computer Science University of Birmingham United Kingdom Department of Computing and Mathematics Manchester Metropolitan University United Kingdom Division of Informatics Imaging and Data Sciences University of Manchester United Kingdom Centre for Computational Imaging and Modelling in Medicine University of Manchester United Kingdom Department of Computing Impeiral College London United Kingdom

Deep learning-based image registration methods have shown state-of-the-art performance and rapid inference speeds. Despite these advances, many existing approaches fall short in capturing spatially varying information in non-local regions of feature maps due to the reliance on spatially-shared convolution kernels. This limitation leads to suboptimal estimation of deformation fields. In this paper, we propose a 3D Spatial-Awareness Convolution Block (SACB) to enhance the spatial information within feature representations. Our SACB estimates the spatial clusters within feature maps by leveraging feature similarity and subsequently parameterizes the adaptive convolution kernels across diverse regions. This adaptive mechanism generates the convolution kernels (weights and biases) tailored to spatial variations, thereby enabling the network to effectively capture spatially varying information. Building on SACB, we introduce a pyramid flow estimator (named SACB-Net) that integrates SACBs to facilitate multi-scale flow composition, particularly addressing large deformations. Experimental results on the brain IXI and LPBA datasets as well as Abdomen CT datasets demonstrate the effectiveness of SACB and the superiority of SACB-Net over the state-of-the-art learning-based registration methods. The code is available at https://***/x-xc/SACB_Net. Copyright © 2025, The Authors. All rights reserved.

关键词： Image registration

Efficient Batch Verification and Privacy-Preserving data Aggregation Scheme in V2G Networks

学校读者我要写书评

暂无评论

IEEE Transactions on Vehicular Technology 2025年

作者： Wang, Congcong Wang, Chen Shen, Jian Vasilakos, Athanasios V. Wang, Bin Wang, Weizheng Department of Software Engineering Nanjing University of Information Science and Technology Nanjing210044 China the Zhejiang Key Laboratory of Digital Fashion and Data Governance the Zhejiang Provincial International Cooperation Base for Science and Technology on Cloud Computing Security and Data Aggregation Zhejiang Sci-Tech University Hangzhou310018 China Zhejiang Provincial Innovation Center of Advanced Textile Technology Shaoxing312000 China College of Computer Science and Information Technology IAU Saudi Arabia Grimstad4879 Norway Network and Data Security Hangzhou310053 China Computer Science Department City University of Hong Kong Hong Kong the Division of Computer Science University of Aizu Fukushima965–8580 Japan

In recent years, against the backdrop of smart grids, vehicle-to-grid (V2G) interaction has become a hot topic. With the increasing number of new energy vehicles, handling massive amounts of electricity transaction data has become an urgent problem in need of resolution. Ensuring the security of the vast quantities of electricity transaction data has emerged as a significant challenge. Therefore, two data aggregation schemes are proposed in this paper, both supporting batch verification and privacy protection in centralized and distributed scenarios, respectively. Firstly, an improved Paillier cryptosystem is utilized to achieve lightweight encryption in the proposed schemes. In addition, a novel distributed data aggregation structure that combines a large number of geographically distributed fog nodes with centralized cloud servers is introduced. This structure not only meets the requirements of lightweight encryption and resistance against single point failures but also reduces the communication complexity of distributed data aggregation to O(n). According to the security proofs and experimental analysis, it can be demonstrated that both proposed schemes are effective and secure, meeting the lightweight requirements of modern smart grid systems. © 2025 IEEE. All rights reserved,

关键词： data aggregation

Back to the Future: Challenges of Sparse and Irregular Medical Image Time Series

学校读者我要写书评

暂无评论

Back to the Future: Challenges of Sparse and Irregular M...

Workshop on Longitudinal Disease Tracking and Modeling with Medical Images and data, LDTM 2024, 5th International Workshop on Multiscale Multimodal Medical Imaging, MMMI 2024, 1st Workshop on Machine Learning for Multimodal/-sensor Healthcare data, ML4MHD2024 and Workshop on Multimodal Learning and Fusion Across Scales for Clinical Decision Support, ML-CDS 2024 held in conjunction with the 27th International Conference on Medical Image computing and Computer Assisted Intervention, MICCAI 2024

作者： Disch, Nico Albert Peretzke, Robin Roy, Saikat Ulrich, Constantin Zimmerer, David Stiefelhagen, Rainer Kleesiek, Jens Maier-Hein, Klaus Division of Medical Image Computing German Cancer Research Center Heidelberg Germany HIDSS4Health - Helmholtz Information and Data Science School for Health Karlsruhe Heidelberg Germany Faculty of Mathematics and Computer Science University of Heidelberg Heidelberg Germany Pattern Analysis and Learning Group Department of Radiation Oncology Heidelberg University Hospital Heidelberg Germany Karlsruhe Institute of Technology Karlsruhe Germany University Hospital Essen Essen Germany University Hospital Essen West German Cancer Center Essen Essen Germany Medical Faculty Heidelberg University of Heidelberg Heidelberg Germany

ISBN: (纸本)9783031845246

In longitudinal medical image analysis, most work focuses on regularly sampled images, or on tasks like regression or classification. However, in the clinical context, images are frequently generated irregularly due to factors such as cost constraints. This work tackles the problem of irregularly sampled longitudinal medical imaging by evaluating methods in segmentation and reconstruction tasks. We examine the reconstruction for real-life MRI from patients with Alzheimer’s Disease (AD), where the model predicts future MRI scans from an arbitrary number of input images sampled at irregular intervals. However, experiments show that most models cannot surpass the performance of using an image from the same patient at a different time point as a baseline. Therefore, we conducted experiments on a synthetic dataset to better isolate effects in temporal learning. These experiments offer insights into model behavior and serve as a benchmark for validating models’ longitudinal learning capability, thus providing a straightforward method to assess temporal understanding in a controlled environment. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Neurodegenerative diseases

A Failure-Free and Efficient Discrete Laplace Distribution for Differential Privacy in MPC

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Tjuawinata, Ivan Wang, Jiabo Yang, Mengmeng Lyu, Shanxiang Wang, Huaxiong Lam, Kwok-Yan Strategic Centre for Research in Privacy-Preserving Technologies & Systems Nanyang Technological University Singapore Information Security and Privacy Group Data61 CSIRO Australia College of Cyber Security Jinan University Guangzhou China Division of Mathematical Sciences Nanyang Technological University Singapore College of Computing and Data Science Nanyang Technological University Singapore

In an MPC-protected distributed computation, although the use of MPC assures data privacy during computation, sensitive information may still be inferred by curious MPC participants from the computation output. This can be observed, for instance, in the inference attacks on either federated learning or a more standard statistical computation with distributed inputs. In this work, we address this output privacy issue by proposing a discrete and bounded Laplace-inspired perturbation mechanism along with a secure realization of this mechanism using MPC. The proposed mechanism strictly adheres to a zero failure probability, overcoming the limitation encountered on other existing bounded and discrete variants of Laplace perturbation. We provide analyses of the proposed differential privacy (DP) perturbation in terms of its privacy and utility. Additionally, we designed MPC protocols to implement this mechanism and presented performance benchmarks based on our experimental setup. The MPC realization of the proposed mechanism exhibits a complexity similar to the state-of-the-art discrete Gaussian mechanism, which can be considered an alternative with comparable efficiency while providing stronger differential privacy guarantee. Moreover, efficiency of the proposed scheme can be further enhanced by performing the noise generation offline while leaving the perturbation phase online. © 2025, CC BY-NC-ND.

关键词： Differential privacy