Bayesian inference of mixed-effects ordinary differential equations models using heavy-tailed distributions

A mixed-effects ordinary differential equation (ODE) model is proposed to describe complex dynamical systems. In order to make the inference of ODE parameters robust against the outlying observations and subjects, a class of heavy-tailed distributions is applied to model the random effects of ODE pa...

Full description

Saved in:
Bibliographic Details
Published inComputational statistics & data analysis Vol. 137; pp. 233 - 246
Main Authors Liu, Baisen, Wang, Liangliang, Nie, Yunlong, Cao, Jiguo
Format Journal Article
LanguageEnglish
Published Elsevier B.V 01.09.2019
Subjects
Bayesian theory
differential equation
Metropolis–Hastings
normal distribution
Outliers
Pharmacokinetics
Scale mixtures of multivariate normal distributions
Smoothing spline
Outliers
Metropolis–Hastings
Scale mixtures of multivariate normal distributions
Pharmacokinetics
Smoothing spline
Online AccessGet full text

Cover

More Information
Summary:A mixed-effects ordinary differential equation (ODE) model is proposed to describe complex dynamical systems. In order to make the inference of ODE parameters robust against the outlying observations and subjects, a class of heavy-tailed distributions is applied to model the random effects of ODE parameters and measurement errors in the data. The heavy-tailed distributions are so flexible that they include the conventional normal distribution as a special case. An MCMC method is proposed to make inferences on ODE parameters within a Bayesian hierarchical framework. The proposed method is demonstrated by estimating a pharmacokinetic mixed-effects ODE model. The finite sample performance of the proposed method is evaluated using some simulation studies.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0167-9473
1872-7352
DOI:10.1016/j.csda.2019.03.001