Missing Data: Dial M for ???
缺失数据：拨 M 找寻什么？AI 解读 ↓
← 返回论文列表与筛选
阅读说明
本页整理公开书目信息，不生成学术质量评分。AI 解读根据题名及所注明的摘要或原文片段生成，未完成全文核验，可能有误；具体假设、证明与结论请核对原文。摘要保留来源文本，未经模型改写。
DOI 首次登记不等于接收或发表日期。部分更正、评论等条目没有独立摘要；缺少可靠材料时保留待补状态。
引用量采用 OpenAlex 单条关联记录，不合计多个记录；关联记录可能包含正式发表版本，不等于此版本单独引用。缺失不视为零，引用指标不代表论文质量。
关注作者按书目中的完整姓名匹配，同名和姓名变体需人工辨别。相关论文先按共享细分方向和具体方法排列，再参考通用方法与发表日期。
分类依据与纠错
AI 根据所提供的论文材料归类：摘要为跨学科缺失数据回顾，提及潜在结果不足以将全文归入因果推断细分。
反馈分类问题正在核对关联版本…
代码、数据与补充材料
正在核对对应版本的资源…
关注这篇论文的作者
缺失数据：拨 M 找寻什么？
- 研究问题
- 如何理解真实缺失数据机制与人为构造的潜变量之间的联系，以及它们在统计建模、计算和科学应用中的作用？
- 方法与思路
- 作者以简短的个人历史回顾串联多个以 M 开头的主题，从 McKendrick 的疫情分析故事及其与 EM 算法的联系展开。
- 材料陈述的贡献
- 摘要强调未观测变量构造在建模与计算中的重要性，并概览缺失机制研究的理论、方法和跨学科实践发展。
- 阅读关注点
- 应核对真实数据缺失与主动引入潜变量的区别、历史故事与 EM 的具体联系，以及各主题的论述范围和引用来源。
- 阅读边界
- 仅据 OpenAlex 索引摘要，稿件版本未注明；这是作者自述的简短个人回顾，历史与引用并不完整，不能视为系统综述。
原文摘要
The question mark is common notation for the missing data that occur in most applied statistical analyses. Over the past century, statisticians and other scientists not only have invented numerous methods for handling missing/incomplete data, but also have invented many forms of missing data, including data augmentation, hidden states, latent variables, potential outcome, and auxiliary variables. Purposely constructing unobserved/unobservable variables offers an extraordinarily flexible and powerful framework for both scientific modeling and computation and is one of the central statistical contributions to natural, engineering, and social sciences. In parallel, much research has been devoted to better understanding and modeling of real-life missing-data mechanisms; that is, the unintended data selection process that prevents us from observing our intended data. This article is a very brief and personal tour of these developments, and thus necessarily has much missing history and citations. The tour consists of a number of M's, starting with a historic story of the mysterious method of McKendrick for analyzing an epidemic study and its link to the EM algorithm, the most popular and powerful method of the twentieth century for fitting models involving missing data and latent variables. The remaining M's touch on theoretical, methodological, and practical aspects of missingdata problems, highlighted with some common applications in social, computational, biological, medical, and physical sciences.