source：Statistics School of SWUFE Release date：2016-04-01 Views：406
Theme：Logistic regression with outcome and covariates missing separately or simultaneously
Speaker：Prof. Shenming Li
Hosted by：Prof. Jinguo Gong
Time： AM10:30-11:30,April 07, 2016
Place： Academic Meeting Room ( B212 in Tongbo Building)
Organizers： School of Statistics; Scientific Bureau
Prof. Shenming Li, a professor of the Department of Statistics in Taiwan's Feng Chia University, used to work as president of School of Finance and Business School, and is currently the Distinguished Professor of Feng Chia University.He has published many articles and sample survey of high-level academic actuarial mathematics in Biometrics, Annals of the Institute of Statistical Mathematics and other world-class academic journals.He has received numerous academic awards, including Taiwan Citation Classic Award in 2002, and a multiple of the NSC of rewards (similar to national research projects). His main research areas are sampling design, actuarial mathematics, biostatistics, statistical ecology and TCM statistics.
We propose estimation methods for fitting logistic regression with outcome and covariate variables missing separately or simultaneously. One of the two proposed estimators is an extension of the validation likelihood estimator of Breslow and Cain (1988). The other is a joint conditional likelihood estimator based on both validation and non-validation data. Large sample properties of the proposed estimators are studied under certain regularity conditions. Simulation results show that the joint conditional likelihood estimator is more efficient than the validation likelihood estimator, weighted estimator, and complete-case estimator. The practical use of the proposed methods is illustrated with data from a cable TV study in Taiwan.