A new on-line approach for AIDS modeling and prediction through dynamic data systems identification (DDSI) method

A preliminary application of the dynamic data system identification (DDSI) method to AIDS modeling is presented. The historical occurrences of AIDS disease are considered as the output of a stochastic dynamic system stimulated by the random inputs. The AIDS modeling and prediction by DDSI is performed for the US nationwide AIDS monthly observed data. One autoregressive moving-average model and another combined model are established. Several features of AIDS data, including the exponential and periodic growth, are explored by analyzing the established model.<>