:::

．單位首頁

．系所簡介

．系主任

．強力師資

．研究所

．e35教學卓越計畫

．課程概況

．各類學程

．期刊論文

．招生資訊

．演講專區

．榮譽榜

．系學會

．逢甲校友資訊網

．系友天地

．就業訊息

96年3月02日(五) 2:00 ~3:00 P.M.

演講日期： 96年3月2日(五) 2:00 ~3:00 P.M.
演講者姓名： 王清雲 教授
演講者服務單位：Fred Hutchingson Cancer Research Center, Washington, U.S.A
演講題目：Logistic Regression with Non-ignorable Missing Covariate Data
摘要：Conditional score; Corrected score; Joint modeling; Local We study longitudinal comorbidity in relation to non-cancer survival. Comorbidity measures contain multiple condition indicators, and a weighted comorbidity index is often applied in survival analysis. For each subject, longitudinal comorbidy measurements may be considered as observed variables from an unobserved longitudinal trajectory along with random errors. While a subject's survival may be associated with the observed comorbidity measurements, the hazard function is more likely to be determined by the long-term underlying comorbidity trajectory. We apply some statistical methods for joint modeling to some Medicare beneficiaries aged 66 and older.

Key words:Missing at random, Non-ignorable missing, Sensitivity Analysis.