统计学每日论文速递[03.19]

访问arxivdaily.com，获取带摘要的学术速递，更多学科、收藏、评论、搜索……应有尽有。同步公众号(arXiv每日学术速递)，欢迎关注

stat 方向，今日共计31篇

【1】 Robust-to-outliers square-root LASSO, simultaneous inference with a MOM approach
标题：稳健的异常值平方根套索，用矩量法同时推断
作者：G. Finocchio,A. Derumigny,K. Proksch
备注：70 pages
链接：https://arxiv.org/abs/2103.10420

【2】 Best estimator for bivariate Poisson regression
标题：二元Poisson回归的最佳估计
作者：André Gillibert,Jacques Bénichou,Bruno Falissard
备注：25 pages, 13 figures
链接：https://arxiv.org/abs/2103.10365

【3】 Probabilistic Forecasting of Regional Net-load with Conditional Extremes and Gridded NWP
标题：具有条件极值和网格化数值预报的区域净负荷概率预测
作者：Jethro Browell,Matteo Fasiolo
链接：https://arxiv.org/abs/2103.10335

【4】 Generalized infinite factorization models
标题：广义无限因式分解模型
作者：Lorenzo Schiavon,Antonio Canale,David B. Dunson
机构*：Duke University, Durham, North Carolina , U.S.A., March
链接：https://arxiv.org/abs/2103.10333

【5】 Confidence Regions Near Singular Information and Boundary Points With Applications to Mixed Models
标题：奇异信息和边界点附近的置信区域及其在混合模型中的应用
作者：Karl Oskar Ekvall,Matteo Bottai
机构*：karl. oskar. ekvallcki.se, matteo. bottaioki. se, Division of Biostatistics, Institute of Environmer Medicine, Karolinska Institute, Nobels vag , Stockholm, Sweden, March
链接：https://arxiv.org/abs/2103.10236

【6】 Identification of Underlying Dynamic System from Noisy Data with Splines
标题：用样条函数从含噪数据中识别潜在动力系统
作者：Yujie Zhao,Xiaoming Huo,Yajun Mei
机构*：Georgia Institute of Technology, March
链接：https://arxiv.org/abs/2103.10231

【7】 Bayesian Imaging With Data-Driven Priors Encoded by Neural Networks: Theory, Methods, and Algorithms
标题：神经网络编码数据驱动先验的贝叶斯成像：理论、方法和算法
作者：Matthew Holden,Marcelo Pereyra,Konstantinos C. Zygalakis
机构*：Heriot-Watt University, Edinburgh, Scotland, University of Edinburgh, Edinburgh, Scotland, Maxwell Institute for Mathematical Sciences Bayes Centre, Potterrow, Edinburgh, Scotland, March
链接：https://arxiv.org/abs/2103.10182

【8】 A Probabilistic State Space Model for Joint Inference from Differential Equations and Data
标题：微分方程与数据联合推理的概率状态空间模型
作者：Jonathan Schmidt,Nicholas Krämer,Philipp Hennig
备注：10 pages (+ 6 pages supplementary material), 6 figures
链接：https://arxiv.org/abs/2103.10153

【9】 Inference and Computation for Sparsely Sampled Random Surfaces
标题：稀疏采样随机曲面的推理与计算
作者：Tomas Masak,Tomas Rubin,Victor Panaretos
机构*：Institut de Mathematiques, Ecole Polytechnique Federale de lausanne
备注：34 pages, 9 figures, 2 appendices
链接：https://arxiv.org/abs/2103.10077

【10】 Bartlett correction of an independence test in a multivariate Poisson model
标题：多元Poisson模型中独立性检验的Bartlett校正
作者：Rolf Larsson
机构*：March
备注：66 pages, 10 figures
链接：https://arxiv.org/abs/2103.10058

【11】 Linear Iterative Feature Embedding: An Ensemble Framework for Interpretable Model
标题：线性迭代特征嵌入：可解释模型的集成框架
作者：Agus Sudjianto,Jinwen Qiu,Miaoqi Li,Jie Chen
机构*：Wells Fargo
链接：https://arxiv.org/abs/2103.09983

【12】 A Statistical Introduction to Template Model Builder: A Flexible Tool for Spatial Modeling
标题：空间建模的灵活工具Template Model Builder的统计介绍
作者：Aaron Osgood-Zimmerman,Jon Wakefield
机构*：In-depth statistical review of Template Model Builder (TMB)method-, ology, Large-scale continuous spatial simulation study assessing the Stochastic, Partial Differential Equations representation of Gaussian Processes in, TMB and R-INLA, Discrete spatial simulation with conditionally constrained spatial dis-, tributions in TMB, Novel nonlinear spatial model to jointly estimate cancer incidence and, mortality in the European Union, Provides example code for fitting popular continuous and discrete spa-, tial models in TMB
备注：84 pages, 30 figures
链接：https://arxiv.org/abs/2103.09929

【13】 Differential analysis in Transcriptomic: The strength of randomly picking 'reference' genes
标题：转录学中的差异分析：随机挑选“参考”基因的力量
作者：Dorota Desaulle,Céline Hoffmann,Bernard Hainque,Yves Rozenholc
机构*：Version , mars, Affiliations D.D. and Y.R. Universite de Paris, UR ,-BioSTM, Biostatistique, Traitement et Modelisation des, donnees biologiques,f-, Paris, France C..h.andb.h. Universite de Paris,cnrs, INSERMUTCBS-, Unite des, technologies Chimiques et Biologiques pour la Sante,f-,o, Paris, France, Contributions Y.R. conceived the iterative use of randomly picked genes to derive a proper differential analysis free, of any knowledge of so-called housekeeping genes. D.D. and Y.R. did the mathematical modeling and analysis of the, procedure. D.D. and Y.R. implemented the procedure to confirm empirically this analysis and derive supplementary, results about the power. They performed the differential analysis on the real data, C.H. and. H. conceived and designed the biological experiments. C.H. performed the biological experiments. C.H., and B. H. commented the results obtained on real data.
备注：30 pages, 2 figures
链接：https://arxiv.org/abs/2103.09872

【14】 Combining Pessimism with Optimism for Robust and Efficient Model-Based Deep Reinforcement Learning
标题：悲观与乐观相结合的稳健高效的基于模型的深度强化学习
作者：Sebastian Curi,Ilija Bogunovic,Andreas Krause
链接：https://arxiv.org/abs/2103.10369

【15】 Optimal soil sampling design based on the maxvol algorithm
标题：基于MAXVOL算法的最优土壤取样设计
作者：Anna Petrovskaia,Gleb Ryzhakov,Ivan Oseledets
机构*：Skolkovo Institute of Science and Technology, Bolshoy Boulevard , bld. , Moscow, www.skoltech. ru, Marchuk Institute of Numerical Mathematics, RAS, Gubkin st , Moscow, Russia , http://www.inm.ras.ru
链接：https://arxiv.org/abs/2103.10337

【16】 Rapidly-converging multigrid reconstruction of cone-beam tomographic data
标题：锥束层析数据的快速收敛多重网格重建
作者：Glenn R. Myers,Andrew M. Kingston,Shane J. Latham,Benoit Recur,Thomas Li,Michael L. Turner,Levi Beeching,Adrian P. Sheppard
机构*：The Australian National University,ac, Australia bInstitut, National de la Sant'e et de la Recherche M'edicale, France
备注：7 pages, 4 figures
链接：https://arxiv.org/abs/2103.10329

【17】 How I failed machine learning in medical imaging -- shortcomings and recommendations
标题：我在医学成像领域的机器学习是如何失败的--缺点和建议
作者：Gaël Varoquaux,Veronika Cheplygina
机构*：INRIA, France, IT University of Copenhagen, Denmark, Disclaimer: this is a working paper, and represents research in progress. For missing references
链接：https://arxiv.org/abs/2103.10292

【18】 Optimal Targeting in Fundraising: A Machine Learning Approach
标题：筹款中的最优目标：一种机器学习方法
作者：Tobias Cagala,Ulrich Glogowsky,Johannes Rincke,Anthony Strittmatter
机构*：March
链接：https://arxiv.org/abs/2103.10251

【19】 Optimal transport framework for efficient prototype selection
标题：用于高效原型选择的最佳运输框架
作者：Karthik S. Gurumoorthy,Pratik Jawanpuria,Bamdev Mishra
链接：https://arxiv.org/abs/2103.10159

【20】 Lossless compression with state space models using bits back coding
标题：使用比特反向编码的状态空间模型的无损压缩
作者：James Townsend,Iain Murray
机构*：University College London, University of Edinburgh
链接：https://arxiv.org/abs/2103.10150

【21】 Robust and Guided Bayesian Reconstruction of Single-Photon 3D Lidar Data: Application to Multispectral and Underwater Imaging
标题：单光子三维激光雷达数据的稳健引导贝叶斯重建：在多光谱和水下成像中的应用
作者：Abderrahim Halimi,Aurora Maccarone,Robert Lamb,Gerald S. Buller,Stephen McLaughlin
备注：10 figures
链接：https://arxiv.org/abs/2103.10122

【22】 Top-m identification for linear bandits
标题：线性土匪的Top-m识别
作者：Clémence Réda,Emilie Kaufmann,Andrée Delahaye-Duriez
机构*：Andree Delahaye-Duriez, Universite de Paris, Inserm UMR , NeuroDidero, F-, Paris, France, clemence. redaCinria. fr, Universite Lille, CNRS, Inria, Centrale Lille, UMR , CRIStAL, F-, Lille, France, Universite Sorbonne Paris Nord, UFR SMBH,F-, Bobigny, France&, Assistance Publique des Hopitaux, de Paris, Hopital Jean Verdier, Service d'Histologie-Embryologie-Cytogenetique, F-, Bondy, France
链接：https://arxiv.org/abs/2103.10070

【23】 Approximation for Probability Distributions by Wasserstein GAN
标题：Wasserstein GAN对概率分布的逼近
作者：Yihang Gao,Michael K. Ng
备注：15 pages
链接：https://arxiv.org/abs/2103.10060

【24】 Probabilistic Simplex Component Analysis
标题：概率单纯形成分分析
作者：Ruiyuan Wu,Wing-Kin Ma,Yuening Li,Anthony Man-Cho So,Nicholas D. Sidiropoulos
机构*：The Chinese University of Hong Kong, Hong Kong SAR of China, University of Virginia, Charlottesville, Virginia , USA, March
链接：https://arxiv.org/abs/2103.10027

【25】 Learning Time Series from Scale Information
标题：从尺度信息中学习时间序列
作者：Yuan Yang,Jie Ding
机构*：Harvard University, Oxford Street, Cambridge, MA , USA , Oxford Street, Cambridge MA , USA
链接：https://arxiv.org/abs/2103.10026

【26】 Decision Theoretic Bootstrapping
标题：决策论自举
作者：Peyman Tavallali,Hamed Hamze Bajgiran,Danial J. Esaid,Houman Owhadi
链接：https://arxiv.org/abs/2103.09982

【27】 Outcome-guided Sparse K-means for Disease Subtype Discovery via Integrating Phenotypic Data with High-dimensional Transcriptomic Data
标题：通过整合表型数据和高维转译数据来发现疾病亚型的结果导向稀疏K-Means
作者：Lingsong Meng,Dorina Avram,George Tseng,Zhiguang Huo
机构*：University of Pittsburgh, Pittsburgh, USA., University of Florida, Gainesville, USA.
链接：https://arxiv.org/abs/2103.09974

【28】 Understanding Generalization in Adversarial Training via the Bias-Variance Decomposition
标题：通过偏差-方差分解理解对抗性训练中的泛化
作者：Yaodong Yu,Zitong Yang,Edgar Dobriban,Jacob Steinhardt,Yi Ma
链接：https://arxiv.org/abs/2103.09947

【29】 Infinite-Horizon Offline Reinforcement Learning with Linear Function Approximation: Curse of Dimensionality and Algorithm
标题：基于线性函数逼近的无限水平离线强化学习：维数灾难及其算法
作者：Lin Chen,Bruno Scherrer,Peter L. Bartlett
链接：https://arxiv.org/abs/2103.09847

【30】 Impact of the error structure on the design and analysis of enzyme kinetic models
标题：误差结构对酶动力学模型设计和分析的影响
作者：Elham Yousefi,Werner G. Müller
机构*：Johannes Kepler University Linz, and, Werner G. Muller, March
链接：https://arxiv.org/abs/2103.09563

【31】 Escaping Saddle Points in Distributed Newton's Method with Communication efficiency and Byzantine Resilience
标题：具有通信效率和拜占庭弹性的分布式牛顿方法的鞍点逃逸
作者：Avishek Ghosh,Raj Kumar Maity,Arya Mazumdar,Kannan Ramchandran
机构*：UC Berkeley, College of Information and Computer Sciences, UMASS Amherst, Data Science Institute, University of California, San Diego
链接：https://arxiv.org/abs/2103.09424

机器翻译，仅供参考；“机构”仅供参考