Abstract

Export is an activity of selling goods to another country. Indonesia’s main export capital is natural wealth. From natural wealth owned, can be produced various kinds of export goods. Goods that can be exported are goods that are in demand and needed by overseas buyers. Indonesian export commodities consist of petroleum and gas (oil and gas) as well as non-oil and gas. This study aims to predict the value of Indonesia’s oil and gas exports with the Network of Artificial Backpropogation. Data obtained from customs documents of the Directorate General of Customs and Excise (PEB and PIB) obtained from the Central Bureau of Statistics with url. The data used are the last 40 years (1975-2014) data which is divided into 4 decades (10 years/decade). From five architectural models 10-2-1, architecture 10-5-1, architecture 10-2-5-1, architecture 10-5-2-1 and architecture 10-10-1 obtained the best 10-5-1 model With epoch 28630, MSE training 0,0010005738, MSE testing 0,0034417506. With this architecture model obtained 90% prediction accuracy for the export value of Indonesia.