Combination of ARMA and BPNN Model to predict Rice Type and Rice Price

Rainfall is one of the most vital factors for rice cultivation. The predicting how much rainfall will be in the future by historical data is the process of rainfall forecasting. Depending on the forecasted rainfall by ARMA model, specific rice types are suggested to cultivate. And then the price of those rice types are being forecasted by BPNN model. This paper introduces combination of effective ARMA and scalable BPNN model for rice types and rice prices, specifying different aspects of view on dataset to achieve the efficiency of the proposed combined model. The proposed combined model is exploited as case study in the application area of Rainfall and Rice Price Data of Pyapon Region in Ayeyarwaddy Division, Republic of the Union of Myanmar. The input neurons to visible layers of the BPNN model are hereby four main factors influenced on rice price and rice production. The proposed combined model proves that the accuracy is more efficient and effective.