替代默认收缩方法可以提高预测精度、校准和覆盖范围:一项方法比较研究。

IF 1.6 3区 医学 Q3 HEALTH CARE SCIENCES & SERVICES
Mark A van de Wiel, Gwenaël Gr Leday, Martijn W Heymans, Erik W van Zwet, Ailko H Zwinderman, Jeroen Hoogland
{"title":"替代默认收缩方法可以提高预测精度、校准和覆盖范围:一项方法比较研究。","authors":"Mark A van de Wiel, Gwenaël Gr Leday, Martijn W Heymans, Erik W van Zwet, Ailko H Zwinderman, Jeroen Hoogland","doi":"10.1177/09622802251338440","DOIUrl":null,"url":null,"abstract":"<p><p>While shrinkage is essential in high-dimensional settings, its use for low-dimensional regression-based prediction has been debated. It reduces variance, often leading to improved prediction accuracy. However, it also inevitably introduces bias, which may harm two other measures of predictive performance: calibration and coverage of confidence intervals. Here, the latter evaluates whether the amount of uncertainty is correctly quantified. Much of the criticism stems from the usage of standard shrinkage methods, such as lasso and ridge with a single, cross-validated penalty. Our aim is to show that readily available alternatives may improve predictive performance, in terms of accuracy, calibration or coverage. We study linear and logistic regression. For linear regression, we use small sample splits of a large, fairly typical epidemiological data set to illustrate that usage of differential ridge penalties for covariate groups may enhance prediction accuracy, while calibration and coverage benefit from additional shrinkage of the penalties. Bayesian hierarchical modeling facilitates the latter, including local shrinkage. In the logistic regression setting, we apply an external simulation to illustrate that local shrinkage may improve calibration with respect to global shrinkage, while providing better prediction accuracy than other solutions, like Firth's correction. The potential benefits of the alternative shrinkage methods are easily accessible via example implementations in R, including the estimation of multiple penalties. A synthetic copy of the large data set is shared for reproducibility.</p>","PeriodicalId":22038,"journal":{"name":"Statistical Methods in Medical Research","volume":" ","pages":"9622802251338440"},"PeriodicalIF":1.6000,"publicationDate":"2025-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Alternatives to default shrinkage methods can improve prediction accuracy, calibration, and coverage: A methods comparison study.\",\"authors\":\"Mark A van de Wiel, Gwenaël Gr Leday, Martijn W Heymans, Erik W van Zwet, Ailko H Zwinderman, Jeroen Hoogland\",\"doi\":\"10.1177/09622802251338440\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><p>While shrinkage is essential in high-dimensional settings, its use for low-dimensional regression-based prediction has been debated. It reduces variance, often leading to improved prediction accuracy. However, it also inevitably introduces bias, which may harm two other measures of predictive performance: calibration and coverage of confidence intervals. Here, the latter evaluates whether the amount of uncertainty is correctly quantified. Much of the criticism stems from the usage of standard shrinkage methods, such as lasso and ridge with a single, cross-validated penalty. Our aim is to show that readily available alternatives may improve predictive performance, in terms of accuracy, calibration or coverage. We study linear and logistic regression. For linear regression, we use small sample splits of a large, fairly typical epidemiological data set to illustrate that usage of differential ridge penalties for covariate groups may enhance prediction accuracy, while calibration and coverage benefit from additional shrinkage of the penalties. Bayesian hierarchical modeling facilitates the latter, including local shrinkage. In the logistic regression setting, we apply an external simulation to illustrate that local shrinkage may improve calibration with respect to global shrinkage, while providing better prediction accuracy than other solutions, like Firth's correction. The potential benefits of the alternative shrinkage methods are easily accessible via example implementations in R, including the estimation of multiple penalties. A synthetic copy of the large data set is shared for reproducibility.</p>\",\"PeriodicalId\":22038,\"journal\":{\"name\":\"Statistical Methods in Medical Research\",\"volume\":\" \",\"pages\":\"9622802251338440\"},\"PeriodicalIF\":1.6000,\"publicationDate\":\"2025-05-29\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Statistical Methods in Medical Research\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1177/09622802251338440\",\"RegionNum\":3,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"HEALTH CARE SCIENCES & SERVICES\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Statistical Methods in Medical Research","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1177/09622802251338440","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"HEALTH CARE SCIENCES & SERVICES","Score":null,"Total":0}
引用次数: 0

摘要

虽然收缩在高维环境中是必不可少的,但它在基于低维回归的预测中的使用一直存在争议。它减少了方差,通常会提高预测的准确性。然而,它也不可避免地引入了偏差,这可能会损害预测性能的其他两个指标:校准和置信区间的覆盖。在这里,后者评估不确定性的数量是否被正确量化。许多批评来自标准收缩方法的使用,如套索和脊与一个单一的,交叉验证的惩罚。我们的目标是表明,在准确性、校准或覆盖范围方面,随时可用的替代方案可能提高预测性能。我们研究线性和逻辑回归。对于线性回归,我们使用一个大的、相当典型的流行病学数据集的小样本分割来说明对协变量组使用差异脊惩罚可以提高预测精度,而校准和覆盖受益于额外的惩罚的缩小。贝叶斯分层建模促进了后者,包括局部收缩。在逻辑回归设置中,我们应用外部模拟来说明局部收缩可以改善相对于全局收缩的校准,同时提供比其他解决方案(如Firth的校正)更好的预测精度。通过R中的示例实现,可以很容易地获得备选收缩方法的潜在好处,包括对多重惩罚的估计。为了再现性,共享大数据集的合成副本。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Alternatives to default shrinkage methods can improve prediction accuracy, calibration, and coverage: A methods comparison study.

While shrinkage is essential in high-dimensional settings, its use for low-dimensional regression-based prediction has been debated. It reduces variance, often leading to improved prediction accuracy. However, it also inevitably introduces bias, which may harm two other measures of predictive performance: calibration and coverage of confidence intervals. Here, the latter evaluates whether the amount of uncertainty is correctly quantified. Much of the criticism stems from the usage of standard shrinkage methods, such as lasso and ridge with a single, cross-validated penalty. Our aim is to show that readily available alternatives may improve predictive performance, in terms of accuracy, calibration or coverage. We study linear and logistic regression. For linear regression, we use small sample splits of a large, fairly typical epidemiological data set to illustrate that usage of differential ridge penalties for covariate groups may enhance prediction accuracy, while calibration and coverage benefit from additional shrinkage of the penalties. Bayesian hierarchical modeling facilitates the latter, including local shrinkage. In the logistic regression setting, we apply an external simulation to illustrate that local shrinkage may improve calibration with respect to global shrinkage, while providing better prediction accuracy than other solutions, like Firth's correction. The potential benefits of the alternative shrinkage methods are easily accessible via example implementations in R, including the estimation of multiple penalties. A synthetic copy of the large data set is shared for reproducibility.

求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
Statistical Methods in Medical Research
Statistical Methods in Medical Research 医学-数学与计算生物学
CiteScore
4.10
自引率
4.30%
发文量
127
审稿时长
>12 weeks
期刊介绍: Statistical Methods in Medical Research is a peer reviewed scholarly journal and is the leading vehicle for articles in all the main areas of medical statistics and an essential reference for all medical statisticians. This unique journal is devoted solely to statistics and medicine and aims to keep professionals abreast of the many powerful statistical techniques now available to the medical profession. This journal is a member of the Committee on Publication Ethics (COPE)
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信