SIMPLICITY VS COMPLEXITY IN DETERMINISTIC MODELS: AN APPLICATION TO AIDS DATA

Publisher Website
Google Scholar
Abstract
This paper addresses the issue of how much complexity can be added to an AIDS model and still be supported by the data. Attention focuses on three models for a particular risk group in a particular geographic location to explain the AIDS epidemic. Statistical analysis shows that the simplest of models perform as well as the more complex ones in explaining the observed patterns in available data. Fitted parameters (risk group size, transmission parameter) are approximately equal for the three models. The models also produce similar predictions for the number of infected and diagnosed AIDS cases. Statistical curvature measures indicate that the simple three-compartment model is sufficiently close to “linear” for statistical inference to be adequate. We conclude that for the purpose of predicting trends of spread, simple models are as appropriate as the more complex ones.

This publication has 0 references indexed in Scilit: