Online Program

Return to main conference page
Saturday, June 1
Computational Statistics
Advances in Regression and Modeling
Sat, Jun 1, 2:45 PM - 3:50 PM
Grand Ballroom J
 

Nonparametric Estimation of a Mixing Distribution for Pharmacokinetic Stochastic Models (305179)

*Alona Kryshchenko, California State University Cannel Islands 
Michael Neely, Laboratory of Applied Pharmacokinetics and Bioinformatics, Children’s Hospital-Los Angeles 
Alan Schumitzky, Department of Mathematics, University of Southern California 
Mike van Guilder, Laboratory of Applied Pharmacokinetics and Bioinformatics, Children’s Hospital-Los Angeles 

Keywords: Mixing distribution, Stochastic models, Nonparametric maximum likelihood, Adaptive grid, Pharmacokinetic population models

In this paper we develop a nonparametric maximum likelihood estimate of the mixing distribution of the parameters of a linear stochastic dynamical system. This includes, for example, pharmacokinetic population models with process and measurement noise that are linear in the state vector, input vector and the process and measurement noise vectors. Most research in mixing distributions only considers measurement noise. The advantages of the models with process noise are that, in addition to the measurements errors, the uncertainties in the model itself are taken into the account. For example, for deterministic pharmacokinetic models, errors in dose amounts, administration times, and timing of blood samples are typically not included. For linear stochastic models, we use linear Kalman-Bucy filtering to calculate the likelihood of the observations and then employ a nonparametric adaptive grid algorithm to find the nonparametric maximum likelihood estimate of the mixing distribution.