Accuracy and reliability in forecasting the inflation rates or predicting it trend correctly is very importance for would be investors, academia, and policy makers. The use of intelligence based model have been found to be invaluable for forecasting financial and economic series like inflation rates exchange rates and stock bond so to mention the few. Researchers have used several parametric models in forecasting exchange rates and other financial and economics data. This paper therefore employs the use of non-parametric approach (artificial neural networks) in forecasting inflation rates. It is an indubitable fact that Artificial Neural networks (ANNs), emulates the information processing capabilities of neurons of the human brain. It uses a distributed representation of the information stored in the networks, and thus resulting in robustness against damage and corresponding fault tolerance. A major advantage of neural networks is their ability to provide flexible mapping between inputs and outputs. The arrangement of the simple units into a multi-layer frame works produces a map between inputs and outputs that is consistent with any underlying functional relationship irrespective of the true functional form. This paper therefore, used three artificial neural networks (Standard Backpropagation (SBP), Scaled Conjugate Gradient (SCG) and Backpropagation based forecasting model for Nigerian and American inflation rates. These models were evaluated using five performance series and a comparison was made with traditional ARIMA models. Inflation rates data of United States of America and Federal Republic of Nigeria were used for empirical illustration. The data were analyzed using both statistical programme for social science (SPSS) and Econometrics view (E-view). The results obtained show that all the ANN models outperformed ARIMA models. The implication of this is that ANN based model can be used to forecast the inflation rates market structure.

Projects

Currently no objects available

Currently no objects available