Paddy a crop that produces rice is a is the most important product for the people of Indonesia. Rice at an affordable price for the community is one of the important factors to maintain national resilience and government stability. In this case, forecasting rice production can be used as a consideration for the government to take a policy because forecasting rice production can provide an overview of information in the form of the number of products produced in the future. This study uses the Backpropagation method to forecast the amount of rice production and to evaluate the error value using the Mean Absolute Percentage Error (MAPE). Forecasting paddy production is carried out in several provinces such as DKI Jakarta, West Java, North Sumatra, Riau, and Banten. This study resulted in the smallest MAPE value with a value of 7.39%. This value is generated from data from the province of West Java with 10 neurons as input neurons, 3 hidden neurons, with initial weight and bias range value from -0.8 to 0.8, a learning rate value of 0.6, and an epoch of 50 times.
Copyrights © 2022