CMStatistics 2020: Start Registration
View Submission - CMStatistics
Title: Semiparametric likelihood inference for heterogeneous survival data under random double truncation Authors:  Achim Doerre - University of Rostock (Germany) [presenting]
Abstract: When survival data are sampled exclusively from already failed units during a fixed data collection period, the lifetimes are subject to random double truncation. We study one important special case of this selective sampling setup, in which units originate from different subgroups with varying lifetime distributions. Point processes are used to model the random emergence of units in the population and sample. We investigate semiparametric likelihood inference to account for selection bias. In addition, we show how information on the pattern of unit emergence can be incorporated to obtain estimators with increased precision. Strategies for alleviating numerical issues and improving performance in the implementation are discussed. We explore the finite-sample properties in a simulation study and demonstrate the suggested approach on a large dataset.