Assume that the kinetics of a drug are described by the nonlinear model shown below:
a) Discuss the relative advantages of weighted least squares and least squares parameter
estimation.
b) Weighted least squares parameter estimation for a model that is nonlinear in the
parameters is more difficult than for models that are linear in the parameters. Discuss
why this is the case.
c) The first criterion with which to judge whether or not a specific system model is a
reliable description of the data is the test on residuals. Discuss why residuals should be
independent, and why weighted residuals should lie in a 21, 1 1 wide band.
d) Discuss the relative merits of least squares, maximum likelihood and Bayesian
approaches to parameter estimation in the context of physiological modeling.
e) Discuss the rationale of optimal experiment design and formalize it