Nonlinear forecasting of hepatitis and AIDS incidence.

The simplex methods of nonlinear forecasting are used to study the data sets of hepatitis A and AIDS in various regions of the United States. The results are compared with those obtained from the traditional ARIMA methods. In many regions, the simplex methods developed from nonlinear dynamical system theory give smaller errors for the data of hepatitis A. A combination of the simplex methods and the traditional ARIMA methods can produce better results for the AIDS data sets.