Copula Models for Dependent Censoring
This chapter provides mathematical infrastructures for copula models, focusing on applications to survival analysis involving dependent censoring. After reviewing the concept of copulas, we introduce measures of dependence, including Kendall’s tau and the cross-ratio function. We also introduce the idea of residual dependence that explains how dependence between event times arises and how it can be modeled by copulas. Finally, we apply copulas for modeling the effect of dependent censoring and analyze the bias of the Cox regression analysis owing to dependent censoring.
KeywordsArchimedean copula Clayton’s copula Cox regression Cross-ratio function Gumbel’s copula Kendall’s tau Residual dependence Univariate Cox regression
- Emura T, Nakatochi M, Matsui S, Michimae H, Rondeau V (2017b) Personalized dynamic prediction of death according to tumour progression and high-dimensional genetic factors: meta-analysis with a joint model. Stat Methods Med Res, https://doi.org/10.1177/0962280216688032
- Emura T, Pan CH (2017). Parametric likelihood inference and goodness-of-fit for dependently left-truncated data, a copula-based approach. Stat Pap, https://doi.org/10.1007/s00362-017-0947-z
- Gumbel EJ (I960). Distributions de valeurs extremes en plusieurs dimensions. PubL Inst Statist. Parids 9: 171–173Google Scholar