Non-parametric fitting of nonlinear equations to experi- mental data without use of initial guessing: a basic computer program

Abstract

In this work, we present a non-parametric method, and the appropriate computer program, for fitting nonlinear multiparametric equations to experimental data. Our method is followed by computation of confidence limits of the parameter estimates. Its performance has been tested on several multiparametric equations, common in the fields of Biochemistry and Biotechnology, and it is a multiparametric expansion of the concept proposed by others for equations having more than two parameters. Good parameter estimates were obtained without a previous knowledge of initial parameter guessing values, and the proposed computer program converges rapidly, in all cases examined within this work.