In the literature, there currently exist a multitude of differential equation based models of both primary and long term HIV infections.
Typically, a nonlinear regression is performed on the model in order to identify disease characteristics of interest. However, there is no
widespread utilization of a mathematically and statistically rigorous methodology for comparing the relative merits of specific models. In
this talk, we will discuss and present results from the application of a statistical significance testing methodology which allows us to compare many of the more popular models to a simpler one. These results suggest which improved fits to data are due to additional mathematical modeling mechanisms (e.g., delays or proliferation terms) and which are simply due to increased degrees of freedom.
Copyright. All Rights Reserved.