An application of Nelder-Mead heuristic-based hybrid algorithms: Estimation of compartment model parameters

Abstract

Compartment models are commonly used tools for nonlinear modeling in pharmacokinetic studies. Parameter estimation of compartment models play a crucial role in drug development. In order to estimate the model parameters, a derivative-based method, called stripping, has been commonly used in drug studies until now. In this study, a derivative free simple local search algorithm, Nelder-Mead Simplex (NMS), is hybridized with two artificial intelligence optimization algorithms, Genetic Algorithm (GA) and Particle Swarm Optimization (PSO). The hybridized algorithms are called GAnMs and PSONMS which are used for parameter estimation. These hybrid algorithms are all population based and do not need any assumptions which make the calculations become easier. Two data sets with two compartment models are preferred as application from the literature. It is seen from the results that the suggested PSONMS is more preferable among the GA, PSO and GANMS with consistence parameter estimates and small error function values. © 2016 CESER PUBLICATIONS.