Prediction Of Coffee Prices With Backpropagation Neural Networks

Abstract

Robusta coffee is a type of coffee plant in Indonesian, sepcialy at west lampung.lampung province have much spices of herbs plant.one of that spices is Robusta Coffe. Robusta coffee is the fourth most reliable source of Robusta coffee in the world. Robusta coffee prices in Lampung are influenced by the price of coffee from Vietnam and Brazil. This research was conducted to determine the predication of coffee prices. the price of this coffee can be used for government reference. This prediction is used for farmers to get more leverage. The training data is data for 2010-2014. Testing data used in 2015-2017. 2018 is used as an output target the algorithm used is backpropagationof neural network. the study uses five architectural models. They are 3-9-1,3-12-1,3-15-1 dan 3-19-1. the best model is3-15-1 with mse value is 0.00904753.therefor, this model is good for predicting prices, especially this result is good for the Lampung

Keywords :Robusta, coffe, prices, prediction, neural network, backpropagation