Price Forecasting by Back Propagation Neural Network Model

The process of predicting what will happen in the future by gathering and analyzing past and current data is referred to as forecasting. When trying to make a good forecasting, Back Propagation Neural Network (BPNN) is constructed with different aspects of viewpoints for the high accuracy of that forecasting. This paper introduces efficient and scalable BPNN model for forecasting, allowing different views on data to fuse the responses of the model in complex and exact forecasting. To exploit the application area of the model, Rice Price Data Set of Pyapon Town in Ayeyarwaddy Division, Republic of the Union of Myanmar was used as case study. Four main factors influenced on rice price and rice production are assumed as input neurons to visible layers of the model. BPNN model with four input factors proves that the accuracy is over 80 percentage.