arXivDaily arXiv每日学术速递 周一至周五更新

高校专区

University of Wisconsin-Madison(威斯康星大学麦迪逊分校)

共收录 71
2602.17894 2026-06-17 stat.ML cs.LG math.ST stat.TH 版本更新

Learning from Biased and Costly Data Sources: Minimax-optimal Data Collection under a Budget

从有偏且昂贵的数据源学习:预算下的极小极大最优数据收集

Michael O. Harding, Vikas Singh, Kirthevasan Kandasamy

机构 * Department of Statistics University of Wisconsin-Madison(统计学系威斯康星大学麦迪逊分校) Department of Biostatistics University of Wisconsin-Madison(生物统计学系威斯康星大学麦迪逊分校) Department of Computer Sciences University of Wisconsin-Madison(计算机科学系威斯康星大学麦迪逊分校)

AI总结 针对预算固定的多源数据收集问题,提出最大化有效样本量的采样方案,结合事后分层估计器,实现极小极大最优风险。

Comments COLT 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.01352 2026-06-17 cs.LG astro-ph.HE astro-ph.IM hep-ex physics.data-an 版本更新

MiniFool -- Physics-Constraint-Aware Minimizer-Based Adversarial Attacks in Deep Neural Networks

MiniFool——深度神经网络中基于物理约束感知的最小化器对抗攻击

Lucie Flek, Oliver Janik, Philipp Alexander Jung, Akbar Karimi, Timo Saala, Alexander Schmidt, Matthias Schott, Philipp Soldin, Matthias Thiesmeyer, Christopher Wiebusch, Ulrich Willemsen

机构 * Bonn-Aachen International Center for Information Technology(波恩-亚琛国际信息科技中心) University of Bonn(波恩大学) Lamarr Institute for Machine Learning and Artificial Intelligence(拉马尔机器学习与人工智能研究所) North Rhine-Westphalia(北莱茵-威斯特法伦州) RWTH Aachen University(亚琛工业大学) III. Physikalisches Institut B(物理研究所B) III. Physikalisches Institut A(物理研究所A) Physikalisches Institut(物理研究所) Erlangen Centre for Astroparticle Physics(埃朗根天体粒子物理中心) Friedrich-Alexander-Universität Erlangen-Nürnberg(埃朗根-纽伦堡弗里德里希-亚历山大大学) Dept. of Physics and Wisconsin IceCube Particle Astrophysics Center(物理系和威斯康星冰立方粒子天体物理学中心) University of Wisconsin—Madison(威斯康星大学麦迪逊分校)

AI总结 提出MiniFool算法,通过最小化结合χ²检验统计量与目标分数偏差的代价函数,生成物理感知的对抗样本,用于测试粒子与天体物理中的神经网络分类器,并量化网络决策的鲁棒性。

Comments Submitted to Computing and Software for Big Science

Journal ref Published in: Eur.Phys.J.C 86 (2026) 6, 641

详情

展开后加载摘要…

URL PDF HTML 收藏
2504.11775 2026-06-16 stat.ML cs.CY cs.LG q-fin.RM 版本更新

Discrimination-free Insurance Pricing with Privatized Sensitive Attributes

基于隐私化敏感属性的无歧视保险定价

Tianhe Zhang, Suhan Liu, Peng Shi

机构 * Department of Risk and Insurance, University of Wisconsin-Madison(风险与保险系,威斯康星大学麦迪逊分校) Department of Statistics and Operations Research, University of North Carolina-Chapel Hill(统计与运筹系,北卡罗来纳大学教堂山分校)

AI总结 针对保险公司无法直接获取敏感属性(如性别、种族)的公平定价问题,提出利用隐私化(加噪)敏感属性估计无歧视保费的方法,并建立理论保证与实证验证。

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.20190 2026-06-11 cs.CV 版本更新

CoVR-R:Reason-Aware Composed Video Retrieval

CoVR-R: 推理感知的组合视频检索

Omkar Thawakar, Dmitry Demidov, Vaishnav Potlapalli, Sai Prasanna Teja Reddy Bogireddy, Viswanatha Reddy Gajjala, Alaa Mostafa Lasheen, Rao Muhammad Anwer, Fahad Khan

机构 * Mohamed bin Zayed University of AI(莫扎德·本·扎耶德人工智能大学) University of Chicago(芝加哥大学) University of Wisconsin-Madison(威斯康星大学麦迪逊分校) Linköping University(林奈大学)

AI总结 提出一种零样本推理优先方法,利用大型多模态模型推断编辑的因果和时序后效,并构建CoVR-Reason基准评估,在隐式效应子集上显著优于强基线。

Comments 9 Pages, 3 Figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2507.11688 2026-06-11 cs.LG 版本更新

Composing Linear Layers from Irreducibles

从不可约元组合线性层

Travis Pence, Daisuke Yamada, Vikas Singh

机构 * University of Wisconsin-Madison(威斯康星大学麦迪逊分校)

AI总结 提出用Clifford代数将线性层分解为双向量(几何基元)的组合,仅需O(log^2 d)参数,在LLM注意力投影中匹配强基线性能。

Comments 35 Pages, 11 Tables, 6 Figures, Appearing in NeurIPS 2025

Journal ref Advances in Neural Information Processing Systems 38 (2025)

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.17717 2026-06-11 cs.AI cs.LG 版本更新

A Survey on Evaluating Quality and Trustworthiness in LLM-Generated Data

评估LLM生成数据的质量与可信度综述

Kaituo Zhang, Mingzhi Hu, Hoang Anh Duy Le, Fariha Kabir Torsha, Zhimeng Jiang, Minh Khai Bui, Chia-Yuan Chang, Yu-Neng Chuang, Zhen Xiong, Ying Lin, Guanchu Wang, Na Zou

机构 * University of Houston(德克萨斯大学休斯敦分校) Worcester Polytechnic Institute(沃思利理工学院) Rice University(里德大学) Texas A&M University(德克萨斯农工大学) University of Wisconsin - Madison(威斯康星大学麦迪逊分校) University of Southern California(南加州大学) University of North Carolina at Charlotte(北卡罗来纳州立大学夏洛特分校)

AI总结 提出LLM数据审计框架,从质量和可信度两个维度系统分类评估指标,分析六种模态数据生成方法的评估缺陷并给出改进建议。

Comments Published at TMLR. Title changed in the final version

Journal ref Transactions on Machine Learning Research, 2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2507.22017 2026-06-10 eess.IV cs.CV 版本更新

Cyst-X: A Multi-Center MRI Benchmark and Federated Learning Framework for Malignancy-Risk Stratification of Pancreatic Cystic Neoplasm

Cyst-X:用于胰腺囊性肿瘤恶性风险分层的多中心MRI基准与联邦学习框架

Hongyi Pan, Gorkem Durak, Elif Keles, Ziliang Hong, Deniz Seyithanoglu, Zheyuan Zhang, Alpay Medetalibeyoglu, Halil Ertugrul Aktas, Andrea Mia Bejar, Yavuz Taktak, Gulbiz Dagoglu Kartal, Mehmet Sukru Erturk, Timurhan Cebeci, Yury Velichko, Lili Zhao, Emil Agarunov, Federica Proietto Salanitri, Concetto Spampinato, Pallavi Tiwari, Ziyue Xu, Sachin Jambawalikar, Ivo G. Schoots, Marco J. Bruno, Chenchan Huang, Candice W. Bolan, Tamas Gonda, Frank H. Miller, Rajesh N. Keswani, Michael B. Wallace, Ulas Bagci

机构 * Machine & Hybrid Intelligence Lab, Department of Radiology, Northwestern University(机器与混合智能实验室,放射科,西北大学) Istanbul Faculty of Medicine, Istanbul University(伊斯坦布尔大学医学学院) Department of Biomedical Engineering and Radiology, University of Wisconsin-Madison(生物医学工程与放射科,威斯康星大学麦迪逊分校) Department of Preventive Medicine, Northwestern University(预防医学系,西北大学) Division of Gastroenterology and Hepatology, New York University(消化内科与肝病科,纽约大学) Department of Electrical, Electronic and Computer Engineering, University of Catania(电气、电子和计算机工程系,卡塔尼亚大学) NVIDIA Department of Radiology, Columbia University(放射科,哥伦比亚大学) Department of Radiology and Nuclear Medicine, Erasmus Medical Center(放射科与核医学科,埃因霍温医学院) Department of Gastroenterology and Hepatology, Erasmus Medical Center(消化内科与肝病科,埃因霍温医学院) Department of Radiology, New York University(放射科,纽约大学) Division of Gastroenterology and Hepatology, Mayo Clinic Florida(消化内科与肝病科,迈阿密诊所佛罗里达分部) Department of Gastroenterology and Hepatology, Northwestern University(消化内科与肝病科,西北大学)

AI总结 提出Cyst-X,一个多中心MRI基准和联邦学习框架,用于IPMN恶性风险分层,结合PanSegNet分割器和3D DenseNet-121分类器,在内部交叉验证中达到0.85的AUC,性能与放射科医生相当。

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.23066 2026-06-09 cs.CV 版本更新

Urban Flood Observations: A hand-labeled training and validation dataset of post-flood inundation

城市洪水观测:一个手标注的训练和验证数据集,用于洪水后淹没区域

Rohit Mukherjee, Hannah K. Friedrich, Beth Tellman, Ariful Islam, Zhijie Zhang, Jonathan Giezendanner, Upmanu Lall, Venkataraman Lakshmi

机构 * Pacific Northwest National Laboratory(太平洋西北国家实验室) University of Arizona(亚利桑那大学) University of Wisconsin–Madison(威斯康星大学麦迪逊分校) Utah State University(犹他州立大学) Massachusetts Institute of Technology(麻省理工学院) Columbia University(哥伦比亚大学) University of Virginia(弗吉尼亚大学)

AI总结 本文提出UFO数据集,用于复杂城市环境中从卫星图像中映射洪水淹没区域,通过手标注数据集验证了分割模型,达到77.3的平均IoU,并评估了两种常用水体产品。

Comments 15 pages, 8 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.05356 2026-06-09 cs.CV cs.LG 版本更新

Mitigating Diffusion Model Hallucinations with Dynamic Guidance

通过动态引导缓解扩散模型幻觉

Kostas Triaridis, Alexandros Graikos, Aggelina Chatziagapi, Grigorios G. Chrysos, Dimitris Samaras

机构 * Stony Brook University(石溪大学) University of Wisconsin-Madison(威斯康星大学麦迪逊分校)

AI总结 针对扩散模型因分数函数过度平滑导致的幻觉问题,提出动态引导方法,沿预定方向选择性锐化分数函数,保留有效语义变化,显著减少幻觉。

Comments Project page: https://cvlab-stonybrook.github.io/DynamicGuidance/

详情

展开后加载摘要…

URL PDF HTML 收藏
2603.20967 2026-06-08 stat.ML cs.LG math.ST stat.TH 版本更新

Hard labels sampled from sparse targets mislead rotation invariant algorithms

从稀疏目标采样的硬标签误导旋转不变算法

Avrajit Ghosh, Bin Yu, Manfred Warmuth, Peter Bartlett

机构 * University of California, Berkeley(加州大学伯克利分校) University of Wisconsin, Madison(威斯康星大学麦迪逊分校)

AI总结 针对稀疏目标下的二分类问题,证明旋转不变算法(如逻辑损失梯度下降)的过风险下界为Ω((d-1)/n),而通过重参数化u_i v_i的非旋转不变算法可实现O(s log d / n)的上界。

Journal ref ICML-2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.05363 2026-06-08 cs.AI 版本更新

MHA-RAG: Improving Efficiency, Accuracy, and Consistency by Encoding Exemplars as Soft Prompts

MHA-RAG:通过将示例编码为软提示来提高效率、准确性和一致性

Abhinav Jain, Xinyu Yao, Thomas Reps, Christopher Jermaine

机构 * Department of Computer Science, Rice University(计算机科学系,里士大学) Department of Computer Science, University of Wisconsin–Madison(计算机科学系,威斯康星大学麦迪逊分校)

AI总结 提出MHA-RAG框架,将领域示例编码为软提示,通过多头注意力机制控制生成,在多个问答基准上相比标准RAG提升20点性能,同时降低10倍推理成本。

Comments 17 pages, 5 figures

详情

展开后加载摘要…

URL PDF HTML 收藏