ABSTRACT
The cornerstone of dynamic modelling in biology, especially data-driven mechanistic modelling, is the estimation of parameters for a given model topology, from experimental data. This is challenging, both in terms of the computation involved, as well as posing the right problem for parameter estimation. This chapter overviews this problem of estimating parameters from data, and also surveys some of the popular algorithms for parameter estimation. As a first step, some simple diagnostics can be performed on the data, to understand their distribution, and determine if there are any obviously erroneous data points. Model identification involves identifying the components of the model, the structure or topology, as well as the kinetics that describe the interactions between the various components. Parameter estimation is at the heart of the entire modelling process, and how well this task is carried out, along with necessary preparatory tasks pre-estimation and diagnostic tasks post-estimation, determines the success of modelling exercise. It can be a linear regression problem.
