Construction of optimal design for nonlinear models involves optimization of certain function of Fisher information matrix which depends on unknown parameter(s) value(s). For a locally optimal design, the unknown parameter(s) are replaced by guess value(s) based on prior knowledge of the experimenter. If the guesses are not close enough to the actual parameter(s) value(s) the resulting design may not be optimal, robust and efficient. To address the problem of constructing inefficient designs based on miss guessed parameter value, we employed a new methodology that identify a subclass of designs with a simple format and restrict consideration to this subclass. A locally D-optimal design for Monod model that is supported at two design point was constructed within this subclass. This approach makes construction of optimal design easier because it specifies the optimum number of support points required for any design in question.
This work by European American Journals is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License