Model Selection for a Chemical Inhibition Process

Abstract

The combined effects of the activities of different chemicals are of interest to many researchers. One chemical may promote or inhibit the effects of a second chemical. The behavior is attributed to the inhibition process and due to the chemical balance for different dose levels of the combining chemicals. Five mathematical models are tested here. Using nonlinear least squares methods we fit two sets of experimental data to all five models and estimate the model parameters. The analysis of the results uses the Akaike's criterion (AIC) and its related techniques to select a best model. The coefficients of variation for the parameter estimates are also used in determining the model selection.