Parameter Estimation for the Compartmental Model

Autor:	Pali Sen, R. C. Srivastava
Rok vydání:	2007
Předmět:	Statistics and Probability education.field_of_study Mathematical optimization Steady state Estimation theory Computation Population Estimator General Medicine Nonlinear system Quadratic form Applied mathematics Point estimation Statistics Probability and Uncertainty education Mathematics
Zdroj:	Biometrical Journal. 30:683-695
ISSN:	1521-4036 0323-3847
DOI:	10.1002/bimj.4710300608
Popis:	An estimation procedure is obtained for a stochastic compartmental model. Compartmental analysis assumes that a system may be divided into homogeneous components, or compartments. The main theory for the compartmental system was studied by Matis and Hartley (1971) with a discrete population in a steady state. All the transitions among the particles are considered to be stochastic in nature. An estimation procedure, Regular Best Asymptotic Normal (RBAN), discussed by Chiang (1956) is investigated for a stochastic m-compartmental system. The detailed proof of the procedure is provided here. Asymptotic properties for the estimator has been studied and computation has been carried out on our proposed nonlinear model. The downhill simplex search method, originally developed by Nelder and Mead (1965), and applied to minimize our quadratic form is inherently nonlinear in nature, thus avoiding the need to evaluate any derivative for point estimation of the parameters. The procedure applied to an experimental situation involving two compartments gives very encouraging results.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::f43c0244044a03f74435b6af35e10ade https://doi.org/10.1002/bimj.4710300608 Zobrazit plný text záznamu Plný text