TY - JOUR
T1 - Classified mixed logistic model prediction
AU - Sun, Hanmei
AU - Nguyen, Thuan
AU - Luan, Yihui
AU - Jiang, Jiming
N1 - Funding Information:
The research of Hanmei Sun and Yihui Luan was supported by the Natural Science Foundation of China Grants 11371227, 61432010 and 11626247. The research of Thuan Nguyen and Jiming Jiang was partially supported by the NSF grants DMS-1509557 and DMS-1512084, respectively. The authors wish to thank Dr. Trisha E. Wong of the Oregon Health & Science University for kindly allow them to use the ECMO data and for helpful discussions. Finally, the authors are grateful to an Associate Editor and two referees for their valuable comments that have led to improvement of the manuscript.
Funding Information:
The research of Hanmei Sun and Yihui Luan was supported by the Natural Science Foundation of China Grants 11371227 , 61432010 and 11626247 . The research of Thuan Nguyen and Jiming Jiang was partially supported by the NSF grants DMS-1509557 and DMS-1512084 , respectively. The authors wish to thank Dr. Trisha E. Wong of the Oregon Health & Science University for kindly allow them to use the ECMO data and for helpful discussions. Finally, the authors are grateful to an Associate Editor and two referees for their valuable comments that have led to improvement of the manuscript.
Publisher Copyright:
© 2018 Elsevier Inc.
PY - 2018/11
Y1 - 2018/11
N2 - We develop a classified mixed logistic model prediction (CMLMP) method for clustered binary data by extending a method proposed by Jiang et al. (2018) for continuous outcome data. By identifying a class, or cluster, that the new observations belong to, we are able to improve the prediction accuracy of a probabilistic mixed effect associated with a future observation over the traditional method of logistic regression and mixed model prediction without matching the class. Furthermore, we develop a new strategy for identifying the class for the new observations by utilizing covariates information, which improves accuracy of the class identification. In addition, we develop a method of obtaining second-order unbiased estimators of the mean squared prediction errors (MSPEs) for CMLMP, which are used to provide measures of uncertainty. We prove consistency of CMLMP, and demonstrate finite-sample performance of CMLMP via simulation studies. Our results show that the proposed CMLMP method outperforms the traditional methods in terms of predictive performance. An application to medical data is discussed.
AB - We develop a classified mixed logistic model prediction (CMLMP) method for clustered binary data by extending a method proposed by Jiang et al. (2018) for continuous outcome data. By identifying a class, or cluster, that the new observations belong to, we are able to improve the prediction accuracy of a probabilistic mixed effect associated with a future observation over the traditional method of logistic regression and mixed model prediction without matching the class. Furthermore, we develop a new strategy for identifying the class for the new observations by utilizing covariates information, which improves accuracy of the class identification. In addition, we develop a method of obtaining second-order unbiased estimators of the mean squared prediction errors (MSPEs) for CMLMP, which are used to provide measures of uncertainty. We prove consistency of CMLMP, and demonstrate finite-sample performance of CMLMP via simulation studies. Our results show that the proposed CMLMP method outperforms the traditional methods in terms of predictive performance. An application to medical data is discussed.
KW - CMLMP
KW - CMMP
KW - Clustered binary data
KW - MSPE
KW - Matching
KW - Mixed logistic model
KW - Mixed model prediction
UR - http://www.scopus.com/inward/record.url?scp=85050077527&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85050077527&partnerID=8YFLogxK
U2 - 10.1016/j.jmva.2018.06.004
DO - 10.1016/j.jmva.2018.06.004
M3 - Article
AN - SCOPUS:85050077527
SN - 0047-259X
VL - 168
SP - 63
EP - 74
JO - Journal of Multivariate Analysis
JF - Journal of Multivariate Analysis
ER -