Stable prediction in high-dimensional linear models

doi:10.1007/s11222-016-9694-6

CORC > 数学与系统科学研究院 > 中国科学院数学与系统科学研究院 > 应用数学研究所

	Stable prediction in high-dimensional linear models
	Lin, Bingqing 1; Wang, Qihua1,2 ; Zhang, Jun 1; Pang, Zhen 3
刊名	STATISTICS AND COMPUTING
	2017-09-01
卷号	27 期号:5 页码:1401-1412
关键词	Model averaging Variable selection Penalized regression Screening
ISSN号	0960-3174
DOI	10.1007/s11222-016-9694-6
英文摘要	We propose a Random Splitting Model Averaging procedure, RSMA, to achieve stable predictions in high-dimensional linear models. The idea is to use split training data to construct and estimate candidate models and use test data to form a second-level data. The second-level data is used to estimate optimal weights for candidate models by quadratic optimization under non-negative constraints. This procedure has three appealing features: (1) RSMA avoids model overfitting, as a result, gives improved prediction accuracy. (2) By adaptively choosing optimal weights, we obtain more stable predictions via averaging over several candidate models. (3) Based on RSMA, a weighted importance index is proposed to rank the predictors to discriminate relevant predictors from irrelevant ones. Simulation studies and a real data analysis demonstrate that RSMA procedure has excellent predictive performance and the associated weighted importance index could well rank the predictors.
资助项目	Natural Science Foundation of Shenzhen University[201542] ; Natural Science Foundation of Shenzhen University[701] ; Natural Science Foundation of Shenzhen University[000360023408] ; National Science Fund for Distinguished Young Scholars in China[10725106] ; National Natural Science Foundation of China[11171331] ; National Natural Science Foundation of China[11331011] ; National Natural Science Foundation of China[11401391] ; Key Lab of Random Complex Structure and Data Science, CAS ; Project of Department of Education of Guangdong Province of China[2014KTSCX112] ; Hong Kong Polytechnic University[G-YBKQ]
WOS研究方向	Computer Science ; Mathematics
语种	英语
出版者	SPRINGER
WOS记录号	WOS:000400831700016
内容类型	期刊论文
源URL	[http://ir.amss.ac.cn/handle/2S8OKBNM/25446]
专题	应用数学研究所
通讯作者	Zhang, Jun
作者单位	1.Shenzhen Univ, Inst Stat Sci, Coll Math & Stat, Shenzhen 518060, Peoples R China 2.Chinese Acad Sci, Acad Math & Syst Sci, Beijing 100190, Peoples R China 3.Hong Kong Polytech Univ, Dept Appl Math, Kowloon, Hong Kong, Peoples R China
推荐引用方式 GB/T 7714	Lin, Bingqing,Wang, Qihua,Zhang, Jun,et al. Stable prediction in high-dimensional linear models[J]. STATISTICS AND COMPUTING,2017,27(5):1401-1412.
APA	Lin, Bingqing,Wang, Qihua,Zhang, Jun,&Pang, Zhen.(2017).Stable prediction in high-dimensional linear models.STATISTICS AND COMPUTING,27(5),1401-1412.
MLA	Lin, Bingqing,et al."Stable prediction in high-dimensional linear models".STATISTICS AND COMPUTING 27.5(2017):1401-1412.