|
Activity Number:
|
235
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 5, 2008 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #301079 |
|
Title:
|
Bayesian Mixture Modeling for Interval-Censored Age Onset of Puberty
|
|
Author(s):
|
Bin Huang and Lili Ding*+ and Siva Sivaganisan
|
|
Companies:
|
Cincinnati Children's Hospital Medical Center and Cincinnati Children's Hospital Medical Center and University Cincinnati
|
|
Address:
|
Center for Epi/Bio, MLC 5041, Cincinnati, OH, 45229,
|
|
Keywords:
|
Bayesian ; Dirichlet Process ; Interval-censored ; Case Study ; Puberty
|
|
Abstract:
|
Age of onset of puberty marks the starting point of the development of secondary sexual characteristics in adolescents. Adolescents who experience onset of puberty at earlier or later age are at risk for adverse physical and psychosocial development. Since subjects could only be seen at predetermined time points, age of onset of puberty is typically interval-censored. Current practice classifies adolescents whose age of onset of puberty falls below 20 or above 80 percentile as early or late timers. We build a Dirichlet process mixture model for the clustering analysis of interval-censored age of onset of puberty, augment interval-censored observations, develop a Gibbs sampling algorithm based on a truncated approximation of the Dirichlet process, and provide techniques to obtain posterior inferences on the number of clusters in the data and classifications of individuals.
|
