REVENUE FORECASTING USING A FEED-FORWARD NEURAL NETWORK AND ARIMA MODEL

41750

Revenue forecasting using intraday updates, which provides managers to make a flexible decisions and plan short-term financing, is a very important problem. In this study, revenue forecasting hybrid model,which is a combination of ARIMA and feed-forward neural network models, is developed. At the end of this study, results with real data sets indicate that the combined model can be an effective way to improve forecasting accuracy achieved by either of the models used separately. This study has been tested in 130 stores of a fashion retail chain. Through this proposed prediction model, the best accuracy of prediction at the end of day could reach up to 80%-85%, and prediction for each hour could reach up to %90-%95.

129

Feed , forward neural networks , ARIMA , revenue forecasting

Sigma Journal Of Engineering and Natural Sciences

134

Sigma Journal Of Engineering and Natural Sciences