Robust Estimation of Location in Matrix Manifolds Using the Projected Frobenius Median

本文提出了一种基于投影弗罗贝尼乌斯中位数的鲁棒位置估计方法,该方法通过在欧氏空间计算中位数并投影到矩阵流形(如施蒂费尔流形、格拉斯曼流形及形状空间),实现了计算高效、具有唯一解、良好鲁棒性及变换等变性的统计推断,并验证了其在模拟数据与真实地震矩张量数据上的有效性。

Houren Hong, Kassel Liam Hingee, Janice L. Scealy, Andrew T. A. WoodMon, 09 Ma📊 stat

Two Localization Strategies for Sequential MCMC Data Assimilation with Applications to Nonlinear Non-Gaussian Geophysical Models

本文提出了一种基于序贯马尔可夫链蒙特卡洛(SMCMC)技术的局部数据同化方案,通过两种利用观测空间稀疏性的新策略,在避免粒子滤波权重退化问题的同时,有效处理了高维非线性非高斯地理物理模型(包括 SWOT 和漂流浮标数据)中的重尾观测噪声,并展示了其优于局部集合变换卡尔曼滤波(LETKF)的性能。

Hamza Ruzayqat, Hristo G. Chipilski, Omar KnioMon, 09 Ma📊 stat

Simultaneously accounting for winner's curse and sample structure in Mendelian randomization: bivariate rerandomized inverse variance weighted estimator

该论文提出了一种双变量重随机化逆方差加权(BRIVW)估计量,通过联合建模 SNP-暴露与 SNP-结局的关联分布并校正样本结构,有效解决了孟德尔随机化中同时存在的赢家诅咒和样本结构偏差问题,从而获得更准确且无偏的因果效应估计。

Xin Liu, Ping Yin, Peng WangMon, 09 Ma📊 stat

Clustering-Based Outcome Models for Clinical Studies: A Scoping Review

这篇综述系统概述了将患者协变量聚类与临床结局模型相结合的方法,区分了利用结局信息构建聚类的“知情”模型与仅基于协变量的“无偏”模型,并探讨了其在高维数据、风险分层及亚组治疗效应估计等临床场景中的应用。

Johannes Vilsmeier, Fabian Eibensteiner, Franz König, Francois Mercier, Robin Ristl, Nigel Stallard, Marc Vandemeulebroecke, Sarah Zohar, Martin PoschMon, 09 Ma📊 stat

Optimizing Complex Health Intervention Packages through the Learn-As-you-GO (LAGO) Design

本文介绍了“边学边做”(LAGO)研究设计,该设计通过在试验过程中分阶段动态优化复杂的多组分健康干预方案,旨在以最小成本实现预设的统计功效和效果目标,从而降低传统随机对照试验在复杂公共卫生干预中失败的风险,并通过 BetterBirth 研究及多项 HIV 与非传染性疾病试验案例论证了其应用价值。

Donna Spiegelman (Center on Methods for Implementation,Prevention Science,,Department of Biostatistics, Yale University), Dong Roman Xu (Southern Medical University Institute for Global Health), Ante Bing (Department of Mathematics,Statistics, Boston University), Guangyu Tong (Section of Cardiovascular Medicine, Department of Internal Medicine, Yale University), Mona Abdo (Center on Methods for Implementation,Prevention Science,,Department of Biostatistics, Yale University), Jingyu Cui (Center on Methods for Implementation,Prevention Science,,Department of Biostatistics, Yale University), Charles Goss (Center for Biostatistics,Data Science, Washington University School of Medicine), John Baptist Kiggundu (Infectious Diseases Research Collaboration), Chris T. Longenecker (Division of Cardiology,Department of Global Health, University of Washington), LaRon Nelson (Yale School of Nursing, Yale University), Drew Cameron (Department of Health Policy,Management, Yale University), Fred Semitala (Infectious Diseases Research Collaboration,,Department of Medicine, Makerere University,,Makerere University Joint AIDS Program), Xin Zhou (Center on Methods for Implementation,Prevention Science,,Department of Biostatistics, Yale University), Judith J. Lok (Department of Mathematics,Statistics, Boston University)Mon, 09 Ma📊 stat

Variable selection in linear mixed model meta-regression with suspected interaction effects -- How can tree-based methods help?

该研究通过真实数据与模拟实验表明,虽然在线性交互效应下传统线性方法表现更优,但在交互效应偏离线性或样本量增加时,基于稳定选择的随机效应树等树基方法能作为线性模型的有力补充,有效解决元回归中交互效应的变量选择问题。

Jan-Bernd Igelmann, Paula Lorenz, Markus PaulyMon, 09 Ma📊 stat

Measuring capacities in multimodal maritime port systems with anchorage queues

本文提出了一种区分“运营容量”与“极限容量”的框架,通过排队论和微分方程模型对多模态港口系统(以休斯顿港为例)进行容量估算,揭示了不同资源在稳定运行与扰动恢复情境下的瓶颈差异,从而为港口规划与韧性分析提供方法支持。

Debojjal Bagchi, Kyle Bathgate, Kenneth N. Mitchell, Magdalena I. Asborno, Marin M. Kress, Stephen D. BoylesFri, 13 Ma📊 stat

Beyond the Oracle Property: Adaptive LASSO in Cointegrating Regressions with Local-to-Unity Regressors

本文推导了自适应 LASSO 估计量在包含局部至单位根回归量的协整回归中的新渐近性质,提出了无需估计局部至单位根或长期协方差参数即可实现一致覆盖的可行置信区间,并通过模拟与实证分析证明其能更准确地刻画有限样本不确定性,克服了传统“神谕性质”在实证应用中覆盖不足的问题。

Karsten Reichold, Ulrike SchneiderFri, 13 Ma📈 econ