Imputed Factor Regression for High-dimensional Block-wise Missing Data

Imputed Factor Regression for High-dimensional Block-wise Missing Data
复制标题

高维分块缺失数据的估算因子回归

DOI:
10.5705/ss.202018.0008
复制
发表时间:
2020
期刊:
影响因子:
1.4
通讯作者:
Annie Qu
Annie Qu
中科院分区:
数学3区
文献类型:
--
作者:
Yanqing Zhang;Niansheng Tang;Annie Qu

文献摘要

参考文献

被引文献

相似文献

在高维生物医学、社会、心理和环境研究中,块丢失数据正变得越来越常见。因此,我们需要有效的降维方法来提取这些数据下的预测的重要信息。现有的降维方法和特征组合对于处理分块丢失的数据是无效的。我们提出了一种针对分块缺失数据的因子模型填充方法,并使用归一化因子回归进行降维和预测。具体地说,我们首先进行筛选以确定重要特征。然后,根据因子模型对这些特征进行输入,并根据输入的特征建立因子回归模型对响应变量进行预测。由于模型的因子结构,该方法利用了所有观测数据的基本信息。此外,即使在块丢失比例很高的情况下,该方法仍然有效。结果表明,在正则性条件下,投入因子回归模型与其预测结果是一致的。我们使用模拟研究将所提出的方法与现有方法进行了比较,然后将其应用于来自《阿尔茨海默病统计杂志:预印本DOI:10.5705/ss.202018.0008》的数据
Block-wise missing data are becoming increasingly common in highdimensional biomedical, social, psychological, and environmental studies. As a result, we need efficient dimension-reduction methods for extracting important information for predictions under such data. Existing dimension-reduction methods and feature combinations are ineffective for handling block-wise missing data. We propose a factor-model imputation approach that targets block-wise missing data, and use an imputed factor regression for the dimension reduction and prediction. Specifically, we first perform screening to identify the important features. Then, we impute these features based on the factor model, and build a factor regression model to predict the response variable based on the imputed features. The proposed method utilizes the essential information from all observed data as a result of the factor structure of the model. Furthermore, the method remains efficient even when the proportion of block-wise missing is high. We show that the imputed factor regression model and its predictions are consistent under regularity conditions. We compare the proposed method with existing approaches using simulation studies, after which we apply it to data from the Alzheimer’s DisStatistica Sinica: Preprint doi:10.5705/ss.202018.0008
DOI: 10.2139/ssrn.2607666
发表时间: 2014-12
期刊: Econometric Modeling: Forecasting eJournal
影响因子: --
作者:
Jianqing Fan;Lingzhou Xue;Jiawei Yao
通讯作者: Jianqing Fan;Lingzhou Xue;Jiawei Yao
DOI: --
发表时间: 2008-12
期刊: --
影响因子: --
作者:
Piyush Rai;Hal Daumé
通讯作者: Piyush Rai;Hal Daumé
DOI: 10.1080/01621459.2016.1261710
发表时间: 2017
影响因子: 3.7
作者:
Zhu H;Shen D;Peng X;Liu LY
通讯作者: Liu LY
在因子模型中预计主成分分析。
DOI: 10.1214/15-aos1364
发表时间: 2016-02
影响因子: 4.5
作者:
Fan J;Liao Y;Wang W
通讯作者: Wang W
影响因子: --
作者:
M. West;J. Nevins;J. Marks;R. Spang;H. Zuzan
通讯作者: M. West;J. Nevins;J. Marks;R. Spang;H. Zuzan