forecasting iranian inflation rates using .st ructura l, time series, and artificial neural networks models

in this paper, i develop three forecasting models: namely structural, times series, and artificial neural networks; to forecast iranian inflation rates. the structural model uses aggregate demand and aggregate supply approach, the time series model is based on the standard arlma technique, and the artificial neural network applies multi-layer back propagation model the latter, which is rooted in physic, cognitive, and computer sciences, mimics human brain to learn any complex pattern and to forecast their future behavior-the results of the forecasting competition show that the back propagation model is able to generate inflation• forecasts much better than the traditional competitors