Work

Bayesian Frailty Models for Correlated Interval-Censored Survival Data

Year: 2010
Type: article
Abstract: Interval-censored time to event data occur in survival analysis when the event time is only known to fall into an interval and these intervals often overlap with each other.Correlated survival data occur when individuals under study are clustered or experience multiple events of interest.For correlated interval-censored data, we study Bayesian parametric frailty models and Bayesian nonparametric frailty models with Dirichlet process mixtures.Statistical analysis and model selection methods based on Monte Carlo simulation are developed.Simulation studies and the analysis of bivariate interval-censored age at onset of puberty illustrate the performance and applications of the proposed methodologies.Γ Gamma function E Expectation Var Variance F Cumulative density function f Probability density function H Cumulative hazard function h Hazard function h 0 Baseline hazard function S Survival function X ∼ LN (µ, σ 2 ) Log-Normal distributed random variable with mean µ and variance σ 2 X ∼ Multi(p) Multinomial distributed random variable with parameters p = (p 1 , . . ., p k ) X ∼ N (µ, σ 2 ) Normal distributed random variable with mean µ and variance σ 2 X ∼ Bernoulli(p) Bernoulli distributed random variable with parameter p X ∼ Beta(α 1 , α 2 ) Beta distributed random variable with parameters α 1 and α 2 X ∼ D(α 1 , α 2 , . . ., α k ) Dirichlet distributed random variable with parameters (α 1 , α 2 , . . ., α k ) X ∼ Exp(λ) Exponential distributed random variable with parameter λ X ∼ G(α, β) Gamma distributed random variable with shape parameter α and inverse scale parameter β X ∼ MVN (0, Σ) Multivariate Normal distributed random vector with mean 0 and variance covariance matrix Σ X ∼ U(a, b) Uniform distributed random variable over the interval [a, b] X ∼ W(γ, α) Weibull distributed random variable with inverse scale parameter γ and shape parameter α a.s.Almost surely BIC Bayesian Information Criteria cdf Cumulative density function CPO Conditional predictive ordinate CRP Chinese restaurant process DAG Direct acyclic graph DPM Dirichlet process mixture iid Independently identically distributed KL Kullback-Leibler LPML Logarithm of pseudomarginal likelihood MCMC Markov chain Monte Carlo MLE Maximum likelihood estimation NGHS The National Heart, Lung, and Blood Institute National Growth and Health Study pdf Probability density function PE Piecewise exponential PH Proportional hazard xi SES Socioeconomic status TDP Truncated Dirichlet process TDPM Truncated Dirichlet process mixture xii
Author: Lili Ding
Language: English
Open Access status: green
OpenAlex ID: W2243015999
FWCI:
Cites: 133
Cited by:
Related to: 10