检索结果-内蒙古大学图书馆

Fine-Grained Partial Label Learning

学校读者我要写书评

暂无评论

Fine-Grained Partial Label Learning

Artificial Intelligence (CAI), IEEE Conference on

作者： Cheng Chen Yueming Lyu Xingrui Yu Jing Li Ivor W Tsang The Australian Artificial Intelligence Institute (AAII) University of Technology Sydney Australia CFAR Agency for Science Technology and Research Singapore IHPC Agency for Science Technology and Research Singapore School of Computer Science and Engineering College of Computing and Data Science Nanyang Technological University Singapore

ISBN: (数字)9798350354096

ISBN: (纸本)9798350354102

Partial Label Learning (PLL) involves associating each instance with a candidate label set, which includes the true label along with false positive labels. Traditional PLL approaches typically assume that these false positive labels follow either a uniform or a non-uniform structure, based on the true label of an instance. However, these assumptions do not fully encompass all aspects of real-world scenarios, especially in cases of finegrained objects with subtle differences in appearance. Therefore, we introduce a fine-grained PLL task specifically designed to handle candidate sets that comprise the true label, partially nonuniform fine-grained labels manifesting in a hierarchical structure, and other false positive labels. Nevertheless, the incorporation of these partially non-uniform, fine-grained labels can result in an intractable posterior, leading to an inconsistent classifier. To address this issue, we propose a Specific Classes-Hierarchical (SCH) regularisation that ensures classifier consistency. Subsequently, the Global Label-Hierarchical-wise Embedding (GLHE) regularisation is introduced, using the refined pseudo labels of positive samples from the obtained consistent classifier. This allows the model to learn more distinctive representations from finegrained instances. Our extensive experiments on datasets such as CIFAR-10, CIFAR-100, and CUB-200 have demonstrated the effectiveness of our approach.

关键词： Complexity theory Phase locked loops Task analysis Artificial intelligence

Accelerating Convergence of Score-Based Diffusion Models, Provably

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Gen Huang, Yu Efimov, Timofey Wei, Yuting Chi, Yuejie Chen, Yuxin Department of Statistics The Chinese University of Hong Kong Hong Kong Department of Statistics and Data Science Wharton School University of Pennsylvania PhiladelphiaPA19104 United States Department of Electrical and Computer Engineering Carnegie Mellon University PittsburghPA15213 United States

Score-based diffusion models, while achieving remarkable empirical performance, often suffer from low sampling speed, due to extensive function evaluations needed during the sampling phase. Despite a flurry of recent activities towards speeding up diffusion generative modeling in practice, theoretical underpinnings for acceleration techniques remain severely limited. In this paper, we design novel training-free algorithms to accelerate popular deterministic (i.e., DDIM) and stochastic (i.e., DDPM) samplers. Our accelerated deterministic sampler converges at a rate O(1/T2 ) with T the number of steps, improving upon the O(1/T ) rate for the DDIM sampler;and our accelerated stochastic sampler converges at a rate O(1/T ), outperforming the rate O(√1/T ) for the DDPM sampler. The design of our algorithms leverages insights from higher-order approximation, and shares similar intuitions as popular high-order ODE solvers like the DPM-Solver-2. Our theory accommodates 2-accurate score estimates, and does not require log-concavity or smoothness on the target distribution. © 2024, CC BY.

关键词： Approximation algorithms

Machine Learning Information Retrieval and Summarisation to Support Systematic Review on Outcomes Based Contracting

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Bilal, Iman Munire Fang, Zheng Arana-Catania, Miguel van Lier, Felix-Anselm Velarde, Juliana Outes Bregazzi, Harry Carter, Eleanor Airoldi, Mara Procter, Rob Department of Computer Science University of Warwick and Alan Turing Institute for Data Science and AI Coventry United Kingdom Faculty of Engineering and Applied Sciences Cranfield University United Kingdom Government Outcomes Lab Blavatnik School of Government University of Oxford Oxford United Kingdom

As academic literature proliferates, traditional review methods are increasingly challenged by the sheer volume and diversity of available research. This article presents a study that aims to address these challenges by enhancing the efficiency and scope of systematic reviews in the social sciences through advanced machine learning (ML) and natural language processing (NLP) tools. In particular, we focus on automating stages within the systematic reviewing process that are time-intensive and repetitive for human annotators and which lend themselves to immediate scalability through tools such as information retrieval and summarisation guided by expert advice. The article concludes with a summary of lessons learnt regarding the integrated approach towards systematic reviews and future directions for improvement, including explainability. © 2024, CC BY.

关键词： Adversarial machine learning

Large Language Models Could Be Rote Learners

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Xu, Yuyang Hu, Renjun Ying, Haochao Wu, Jian Shi, Xing Lin, Wei College of Computer Science and Technology Zhejiang University China School of Data Science and Engineering East China Normal University China State Key Laboratory of Transvascular Implantation Devices The Second Affiliated Hospital Zhejiang University School of Medicine China School of Public Health Zhejiang University China Zhejiang Key Laboratory of Medical Imaging Artificial Intelligence China Alibaba Cloud Computing

Multiple-choice question (MCQ) benchmarks are widely used for evaluating Large Language Models (LLMs), yet their reliability is undermined by benchmark contamination. In this study, we reframe contamination as an inherent aspect of learning and seek to disentangle genuine capability acquisition from superficial memorization in LLM evaluation. First, by analyzing model performance under different memorization conditions, we uncover a counterintuitive trend: LLMs perform worse on memorized MCQs than on non-memorized ones, indicating the coexistence of two distinct learning phenomena, i.e., rote memorization and genuine capability learning. To disentangle them, we propose TrinEval, a novel evaluation framework that reformulates MCQs into an alternative trinity format, reducing memorization while preserving knowledge assessment. Experiments validate TrinEval’s effectiveness in reformulation, and its evaluation reveals that common LLMs may memorize by rote 20.5% of knowledge points (in MMLU on average). Copyright © 2025, The Authors. All rights reserved.

关键词： Teaching

RSRP Measurement Based Channel Autocorrelation Estimation for IRS-Aided Wideband Communication

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Sun, He Zhu, Lipeng Mei, Weidong Zhang, Rui Department of Electrical and Computer Engineering National University of Singapore Singapore117583 Singapore School of Science and Engineering Shenzhen Research Institute of Big Data China The Chinese University of Hong Kong Guangdong Shenzhen518172 China National Key Laboratory of Wireless on Communications University of Electronic Science and Technology of China 611731 China

The passive and frequency-flat reflection of IRS, as well as the high-dimensional IRS-reflected channels, have posed significant challenges for efficient IRS channel estimation, especially in wideband communication systems with significant multi-path channel delay spread. To address these challenges, we propose a novel neural network (NN)-empowered framework for IRS channel autocorrelation matrix estimation in wideband orthogonal frequency division multiplexing (OFDM) systems. This framework relies only on the easily accessible reference signal received power (RSRP) measurements at users in existing wideband communication systems, without requiring additional pilot transmission. Based on the estimates of channel autocorrelation matrix, the passive reflection of IRS is optimized to maximize the average user received signal-to-noise ratio (SNR) over all subcarriers in the OFDM system. Numerical results verify that the proposed algorithm significantly outperforms existing power-measurement-based IRS reflection designs in wideband channels. Copyright © 2024, The Authors. All rights reserved.

关键词： Channel estimation

RETIA: Relation-Entity Twin-Interact Aggregation for Temporal Knowledge Graph Extrapolation

学校读者我要写书评

暂无评论

RETIA: Relation-Entity Twin-Interact Aggregation for Tempora...

International Conference on data engineering

作者： Kangzheng Liu Feng Zhao Guandong Xu Xianzhi Wang Hai Jin National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China Data Science and Machine Intelligence Lab University of Technology Sydney Sydney Australia

Temporal knowledge graph (TKG) extrapolation aims to predict future unknown events (facts) based on historical information, and has attracted considerable attention due to its great practical significance. Accurate representations (embeddings) of entities and relations form the basis of TKG extrapolation. Recent work has been devoted to improving the rationality of entity representations. However, on the one hand, ignoring relation modeling results in incomplete relation representations; therefore, some approaches aggregate only immediately adjacent entities of relations, but this can lead to the "message islands" problem of relation modeling. On the other hand, ignoring the association constraints between relations and entities can make the embeddings of both relations and entities prone to overfitting. To address the abovementioned challenges, we propose an advanced method, namely, RETIA. For the former issue, we generate twin hyperrelation subgraphs for each historical subgraph and then aggregate both the adjacent entities and relations in the hyperrelation subgraphs through a graph convolutional network (GCN). About the latter concern, we propose a twin-interact module (TIM), which provides communication channels for relation aggregation and entity aggregation during the evolution of the historical sequence. Experiments conducted on five public datasets show that RETIA has made great improvements across several evaluation metrics. Our released code is available at https://***/CGCL-codes/RETIA.

关键词：

Joint Routing and Scheduling for In-Vehicle Networks: A Deterministic Perspective

学校读者我要写书评

暂无评论

IEEE Transactions on Intelligent Vehicles

IEEE Transactions on Intelligent Vehicles 2024年 1-15页

作者： Wang, Wenxun Wang, Wenyong Yu, Haisheng Wu, Bochun Zou, Sai Ni, Wei Zhang, Ji School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu China International Institute of Next Generation Internet Macau University of Science and Technology Macau China State Key Laboratory of Integrated Chips and Systems Fudan University Shanghai China College of Big Data and Information Engineering Guizhou University Guiyang China Data61 Business Unit CSIRO Canberra ACT Australia PA Company SC Branch China Telecom Co. LTD Chendu China

Deterministic transmission refers to the guaranteed delivery of data to its destination within a specific time frame along an optimal route, making it an indispensable prerequisite for the operation of autonomous vehicles. The in-vehicle networks (IVNs) face significant challenges in the deterministic transmission of large volumes of perception and control data from and to various vehicle components. In this paper, the cooperative scheduling and routing problem among multi-area control units (ACUs) in IVN is discussed from the view of deterministic transmission, and a ResVmix method is proposed to find the optimal two-layer assignment with ACU cooperation. The multi-cycle queuing and forwarding (multi-CQF) mechanism is employed to schedule the perceived data according to different quality of service (QoS) requirements. The cooperation among multiple ACUs enables deterministic routing of data. The two-layer optimization problems on multi-CQF and ACU are formulated as a+ decentralized partially observable Markov decision processes (Dec-POMDP) and solved using the enhanced multi-agent deep reinforcement learning (MADRL) method. Simulation results demonstrate that ResVmix significantly enhances the deterministic transmission performance of IVN. Compared to traditional shortest path algorithms and state-of-the-art MADRL methods, ResVmix achieves a 22.4% increase in arrival rate and a 20.1% increase in confirmed arrivals, respectively. IEEE

关键词： Reinforcement learning

LLM-Lasso: A Robust Framework for Domain-Informed Feature Selection and Regularization

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Zhang, Erica Goto, Ryunosuke Sagan, Naomi Mutter, Jurik Phillips, Nick Alizadeh, Ash Lee, Kangwook Blanchet, Jose Pilanci, Mert Tibshirani, Robert Department of Management Science and Engineering Stanford University Stanford United States Department of Biomedical Data Science Stanford University School of Medicine Stanford United States Department of Electrical Engineering Stanford University Stanford United States Divisions of Oncology and Hematology Stanford University School of Medicine Stanford United States Department of Electrical and Computer Engineering University of Wisconsin-Madison Madison United States Department of Statistics Stanford University Stanford United States

We introduce LLM-Lasso, a novel framework that leverages large language models (LLMs) to guide feature selection in Lasso 1 regression. Unlike traditional methods that rely solely on numerical data, LLM-Lasso incorporates domain-specific knowledge extracted from natural language, enhanced through a retrieval-augmented generation (RAG) pipeline, to seamlessly integrate data-driven modeling with contextual insights. Specifically, the LLM generates penalty factors for each feature, which are converted into weights for the Lasso penalty using a simple, tunable model. Features identified as more relevant by the LLM receive lower penalties, increasing their likelihood of being retained in the final model, while less relevant features are assigned higher penalties, reducing their influence. Importantly, LLM-Lasso has an internal validation step that determines how much to trust the contextual knowledge in our prediction pipeline. Hence it addresses key challenges in robustness, making it suitable for mitigating potential inaccuracies or hallucinations from the LLM. In various biomedical case studies, LLM-Lasso outperforms standard Lasso and existing feature selection baselines, all while ensuring the LLM operates without prior access to the datasets. To our knowledge, this is the first approach to effectively integrate conventional feature selection techniques directly with LLM-based domain-specific reasoning. © 2025, CC BY-NC-ND.

关键词： Feature Selection

Modeling and Risk Analysis of Cooperative Adaptive Cruise Control Systems Based on Petri Nets and Distributed Edge Intelligence

学校读者我要写书评

暂无评论

IEEE Internet of Things Journal 2025年

作者： Yu, Wangyang Cheng, Yumeng Fang, Xianwen Zhai, Xiaojun Wang, Yinglong Jing, Hongyuan Shaanxi Normal University Chang'an District Key Laboratory of Intelligent Computing and Service Technology for Folk Song Ministry of Culture and Tourism Xi'an710119 China Shaanxi Normal University School of Computer Science Chang'an District Xi'an710119 China Anhui University of Science and Technology Anhui Province Engineering Laboratory for Big Data Analysis and Early Warning Technology of Coal Mine Safety Anhui232001 China University of Essex School of Computer Science and Electronic Engineering Colchester United Kingdom Beijing Union University College of Robotics Beijing100101 China

Fueled by advancements in intelligent transportation systems, the Internet of Vehicles (IoV) seeks to connect smart vehicles, road infrastructure, and users into a unified network, enhancing traffic efficiency and reducing accident risks. Centralized cloud data collection raises concerns about privacy and communication overhead. To address these, Distributed Edge Intelligence (DEI) reduces transmission costs and improves privacy by implementing machine learning at the network edge. In this context, Cooperative Adaptive Cruise Control (CACC) systems, combined with DEI in the IoV framework, enhance transportation system intelligence through real-time data processing and decentralized decision-making. This paper proposes a modeling and analysis method for CACC systems based on Petri nets. The datasets are automatically generated using tools developed by our team, and machine learning methods are utilized to perform risk prediction analysis on the CACC model. From the perspective of Petri nets synchronization, we propose risk mitigation strategies from a design standpoint. The research results show that the proposed method significantly reduces signal accumulation and enhances synchronization in CACC systems. This improvement provides new theoretical support and technical guidance for the design and implementation of CACC systems, ultimately enhancing their safety and reliability. © 2025 IEEE.

关键词： Adaptive cruise control