arXiv - STAT - Other Statistics最新文献_第3页

Interpret the estimand framework from a causal inference perspective 从因果推理的角度解释估计值框架

arXiv - STAT - Other Statistics Pub Date : 2024-06-29 DOI: arxiv-2407.00292

Jinghong Zeng

{"title":"Interpret the estimand framework from a causal inference perspective","authors":"Jinghong Zeng","doi":"arxiv-2407.00292","DOIUrl":"https://doi.org/arxiv-2407.00292","url":null,"abstract":"The estimand framework proposed by ICH in 2017 has brought fundamental\u0000changes in the pharmaceutical industry. It clearly describes how a treatment\u0000effect in a clinical question should be precisely defined and estimated,\u0000through attributes including treatments, endpoints and intercurrent events.\u0000However, ideas around the estimand framework are commonly in text, and\u0000different interpretations on this framework may exist. This article aims to\u0000interpret the estimand framework through its underlying theories, the causal\u0000inference framework based on potential outcomes. The statistical origin and\u0000formula of an estimand is given through the causal inference framework, with\u0000all attributes translated into statistical terms. How five strategies proposed\u0000by ICH to analyze intercurrent events are incorporated in the statistical\u0000formula of an estimand is described, and a new strategy to analyze intercurrent\u0000events is also suggested. The roles of target populations and analysis sets in\u0000the estimand framework are compared and discussed based on the statistical\u0000formula of an estimand. This article recommends continuing study of causal\u0000inference theories behind the estimand framework and improving the estimand\u0000framework with greater methodological comprehensibility and availability.","PeriodicalId":501323,"journal":{"name":"arXiv - STAT - Other Statistics","volume":"47 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141507880","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Fractal dimension, and the problems traps of its estimation 分形维度及其估算的问题陷阱

arXiv - STAT - Other Statistics Pub Date : 2024-06-27 DOI: arxiv-2406.19885

Carlos Sevcik

引用次数: 0

Transfer Learning for High Dimensional Robust Regression 高维鲁棒回归的迁移学习

arXiv - STAT - Other Statistics Pub Date : 2024-06-25 DOI: arxiv-2406.17567

Xiaohui Yuan, Shujie Ren

引用次数: 0

Active Learning for Fair and Stable Online Allocations 主动学习，实现公平稳定的在线分配

arXiv - STAT - Other Statistics Pub Date : 2024-06-20 DOI: arxiv-2406.14784

Riddhiman Bhattacharya, Thanh Nguyen, Will Wei Sun, Mohit Tawarmalani

引用次数: 0

Stock Volume Forecasting with Advanced Information by Conditional Variational Auto-Encoder 用条件变异自动编码器利用高级信息预测股票成交量

arXiv - STAT - Other Statistics Pub Date : 2024-06-19 DOI: arxiv-2406.19414

Parley R Yang, Alexander Y Shestopaloff

引用次数: 0

Producing treatment hierarchies in network meta-analysis using probabilistic models and treatment-choice criteria 利用概率模型和治疗选择标准在网络荟萃分析中建立治疗分层体系

arXiv - STAT - Other Statistics Pub Date : 2024-06-15 DOI: arxiv-2406.10612

Theodoros Evrenoglou, Adriani Nikolakopoulou, Guido Schwarzer, Gerta Rücker, Anna Chaimani

{"title":"Producing treatment hierarchies in network meta-analysis using probabilistic models and treatment-choice criteria","authors":"Theodoros Evrenoglou, Adriani Nikolakopoulou, Guido Schwarzer, Gerta Rücker, Anna Chaimani","doi":"arxiv-2406.10612","DOIUrl":"https://doi.org/arxiv-2406.10612","url":null,"abstract":"A key output of network meta-analysis (NMA) is the relative ranking of the\u0000treatments; nevertheless, it has attracted a lot of criticism. This is mainly\u0000due to the fact that ranking is an influential output and prone to\u0000over-interpretations even when relative effects imply small differences between\u0000treatments. To date, common ranking methods rely on metrics that lack a\u0000straightforward interpretation, while it is still unclear how to measure their\u0000uncertainty. We introduce a novel framework for estimating treatment\u0000hierarchies in NMA. At first, we formulate a mathematical expression that\u0000defines a treatment choice criterion (TCC) based on clinically important\u0000values. This TCC is applied to the study treatment effects to generate paired\u0000data indicating treatment preferences or ties. Then, we synthesize the paired\u0000data across studies using an extension of the so-called \"Bradley-Terry\" model.\u0000We assign to each treatment a latent variable interpreted as the treatment\u0000\"ability\" and we estimate the ability parameters within a regression model.\u0000Higher ability estimates correspond to higher positions in the final ranking.\u0000We further extend our model to adjust for covariates that may affect treatment\u0000selection. We illustrate the proposed approach and compare it with alternatives\u0000in two datasets: a network comparing 18 antidepressants for major depression\u0000and a network comparing 6 antihypertensives for the incidence of diabetes. Our\u0000approach provides a robust and interpretable treatment hierarchy which accounts\u0000for clinically important values and is presented alongside with uncertainty\u0000measures. Overall, the proposed framework offers a novel approach for ranking\u0000in NMA based on concrete criteria and preserves from over-interpretation of\u0000unimportant differences between treatments.","PeriodicalId":501323,"journal":{"name":"arXiv - STAT - Other Statistics","volume":"9 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141507882","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Advances in Machine Learning, Statistical Methods, and AI for Single-Cell RNA Annotation Using Raw Count Matrices in scRNA-seq Data 利用 scRNA-seq 数据中的原始计数矩阵进行单细胞 RNA 注释的机器学习、统计方法和人工智能研究进展

arXiv - STAT - Other Statistics Pub Date : 2024-06-07 DOI: arxiv-2406.05258

Megha Patel, Nimish Magre, Himanshi Motwani, Nik Bear Brown

引用次数: 0

Bayesian Nonparametric Quasi Likelihood 贝叶斯非参数准似然法

arXiv - STAT - Other Statistics Pub Date : 2024-05-31 DOI: arxiv-2405.20601

Antonio R. Linero

{"title":"Bayesian Nonparametric Quasi Likelihood","authors":"Antonio R. Linero","doi":"arxiv-2405.20601","DOIUrl":"https://doi.org/arxiv-2405.20601","url":null,"abstract":"A recent trend in Bayesian research has been revisiting generalizations of\u0000the likelihood that enable Bayesian inference without requiring the\u0000specification of a model for the data generating mechanism. This paper focuses\u0000on a Bayesian nonparametric extension of Wedderburn's quasi-likelihood, using\u0000Bayesian additive regression trees to model the mean function. Here, the\u0000analyst posits only a structural relationship between the mean and variance of\u0000the outcome. We show that this approach provides a unified, computationally\u0000efficient, framework for extending Bayesian decision tree ensembles to many new\u0000settings, including simplex-valued and heavily heteroskedastic data. We also\u0000introduce Bayesian strategies for inferring the dispersion parameter of the\u0000quasi-likelihood, a task which is complicated by the fact that the\u0000quasi-likelihood itself does not contain information about this parameter;\u0000despite these challenges, we are able to inject updates for the dispersion\u0000parameter into a Markov chain Monte Carlo inference scheme in a way that, in\u0000the parametric setting, leads to a Bernstein-von Mises result for the\u0000stationary distribution of the resulting Markov chain. We illustrate the\u0000utility of our approach on a variety of both synthetic and non-synthetic\u0000datasets.","PeriodicalId":501323,"journal":{"name":"arXiv - STAT - Other Statistics","volume":"41 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141258299","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Differentially Private Boxplots 差异化私人方框图

arXiv - STAT - Other Statistics Pub Date : 2024-05-30 DOI: arxiv-2405.20415

Kelly Ramsay, Jairo Diaz-Rodriguez

引用次数: 0

Scaling up archival text analysis with the blockmodeling of n-gram networks -- A case study of Bulgaria's representation in the Osservatore Romano (January-May 1877) 利用 n-gram 网络的块模型扩大档案文本分析 -- 保加利亚在《罗马汇报》（1877 年 1 月至 5 月）中的代表性案例研究

arXiv - STAT - Other Statistics Pub Date : 2024-05-30 DOI: arxiv-2405.20156

Fabio Ashtar Telarico

引用次数: 0