<![CDATA[Coffee is one of Indonesia's essential export commodities and a foreign exchange source for the country. One crucial factor in coffee production development is the planted land area. Therefore, the availability of land for coffee plants in Indonesia needs to be maintained for the continuity of coffee production today and in the future. This study aimed to see the potential for the widespread development of coffee plants on the island of Sumatra. This is because the island of Sumatra is the largest coffee producer in Indonesia, so information about the potential for the development of this plant area needs to be known as early as possible, especially for the agriculture/plantation service and for coffee farmers, so that coffee production can be maintained. The algorithm proposed in this study is the Broyden Fletcher Goldfarb Shanno (BFGS) Quasi-Newton algorithm which can be used to solve data prediction (forecasting) problems. This study uses a dataset of coffee plant areas sourced from the Directorate General of Plantations for 2012-2021. This study was analyzed using 3 (three) network architecture models (4-9-1, 4-18-1, and 4-27-1). Based on the analysis, the results obtained from model 4-18-1 as the best architecture with 100% accuracy with minor MSE testing, which is 0.00036764820. Meanwhile, based on predictions made using the best architecture (predictions for 2022 and 2023), the area of coffee plantations has decreased slightly. So this needs serious attention from the respective provincial governments.]]>
Copyrights © 2023
