Fitting Nonlinear Models to Biological Data by Marquardt's Algorithm

Abstract

Equations describing biological phenomena frequently have parameters entering nonlinearly. A highly efficient method for determining the values of parameters in such equations that will give the best least—squares fit to a set of data is Marquardt's algorithm. A step—by—step derivation of the algorithm is given in terms that require only a knowledge of elementary calculus and matrix algebra. As an example, the method is used to fit the logistic equation to a set of data.