Peramalan Volume Ekspor Crude Palm Oil (CPO) Indonesia Menggunakan Metode Feed Forward Neural Network (FFNN)

Hati, Puspa Damai Kukuh (2026) Peramalan Volume Ekspor Crude Palm Oil (CPO) Indonesia Menggunakan Metode Feed Forward Neural Network (FFNN). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2043221059-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Crude Palm Oil (CPO) merupakan salah satu komoditas ekspor unggulan Indonesia yang berperan penting dalam perekonomian nasional dan perdagangan internasional. Sebagai salah satu eksportir CPO terbesar di dunia, Indonesia perlu menjaga keberlanjutan kinerja ekspornya. Namun, volume ekspor CPO Indonesia menunjukkan pola yang berfluktuasi dari waktu ke waktu dan diduga dipengaruhi oleh perubahan harga internasional CPO. Oleh karena itu, diperlukan suatu metode peramalan yang mampu mempelajari hubungan antara harga internasional CPO sebagai variabel input dan volume ekspor CPO Indonesia sebagai variabel output. Penelitian ini dilakukan menggunakan metode Feed Forward Neural Network (FFNN) dengan algoritma Backpropagation. Data yang digunakan berupa data bulanan harga internasional CPO dan volume ekspor CPO Indonesia periode Januari 2012 hingga Desember 2024. Pemilihan model dilakukan melalui eksplorasi kombinasi neuron pada input layer dan hidden layer menggunakan proporsi data pelatihan dan pengujian sebesar 80:20. Hasil penelitian menunjukkan bahwa model terbaik diperoleh pada arsitektur 4–1–1 dengan learning rate 0,7 dan maximum epoch 65, yang menghasilkan nilai MAPE sebesar 40,8984% dan RMSE sebesar 177.871,78. Berdasarkan model tersebut, volume ekspor CPO Indonesia periode Januari 2025 hingga April 2026 diperkirakan tetap berfluktuasi mengikuti perubahan harga internasional CPO, dengan volume ekspor tertinggi diperkirakan terjadi pada Juni 2025 sebesar 368.908 ton, sedangkan volume terendah diperkirakan terjadi pada Januari 2025 sebesar 81.306 ton. Hasil penelitian menunjukkan bahwa metode ini mampu digunakan untuk memodelkan hubungan antara kedua variabel serta menghasilkan peramalan dengan tingkat akurasi yang cukup (reasonable forecasting).
================================================================================================================================
The Crude Palm Oil (CPO) is one of Indonesia's leading export commodities that plays an important role in the national economy and international trade. As one of the world's largest CPO exporters, Indonesia needs to maintain the sustainability of its export performance. However, the export volume of Indonesian CPO has fluctuated over time and is presumed to be influenced by changes in international CPO prices. Therefore, a forecasting method capable of learning the relationship between international CPO prices as the input variable and Indonesian CPO export volume as the output variable is required. This study employed the Feed Forward Neural Network (FFNN) method with the Backpropagation algorithm. The data consisted of monthly international CPO prices and Indonesian CPO export volume from January 2012 to December 2024. Model selection was conducted by exploring various combinations of neurons in the input and hidden layers using an 80:20 training-testing data split. The results showed that the best model was obtained using a 4-1-1 network architecture with a learning rate of 0.7 and a maximum of 65 epochs, resulting in a MAPE of 40,8984% and an RMSE of 177.871,78. Based on this model, the export volume of Indonesian CPO from January 2025 to April 2026 is forecasted to continue fluctuating in response to changes in international CPO prices, with the highest export volume predicted in June 2025 at 382.454 tons and the lowest in January 2025 at 80.332 tons. The findings indicate that the proposed method is capable of modeling the relationship between the two variables and producing forecasts with reasonable forecasting accuracy.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Backpropagation, Crude Palm Oil (CPO), Feed Forward Neural Network (FFNN), Indonesia's CPO Export Volume, International CPO Price, Backpropagation, Crude Palm Oil (CPO), Feed Forward Neural Network (FFNN), Harga Internasional CPO, Volume Ekspor CPO Indonesia
Subjects:	H Social Sciences > HA Statistics > HA30.3 Time-series analysis
Divisions:	Faculty of Vocational > 49501-Business Statistics
Depositing User:	Puspa Damai Kukuh Hati
Date Deposited:	31 Jul 2026 03:58
Last Modified:	31 Jul 2026 03:58
URI:	http://repository.its.ac.id/id/eprint/140627

Actions (login required)

View Item