Please use this identifier to cite or link to this item: http://repository.ipb.ac.id/handle/123456789/121175
Full metadata record
DC FieldValueLanguage
dc.contributor.advisorRizki, Akbar-
dc.contributor.advisorMasjkur, Mohammad-
dc.contributor.authorUskono, Elisabeth Kurniasari Marcela-
dc.date.accessioned2023-07-09T00:35:44Z-
dc.date.available2023-07-09T00:35:44Z-
dc.date.issued2023-
dc.identifier.urihttp://repository.ipb.ac.id/handle/123456789/121175-
dc.description.abstractMinyak sawit mentah atau Crude Palm Oil (CPO) adalah salah satu komoditas unggulan ekspor Indonesia di sektor pertanian. Indonesia juga merupakan produsen CPO terbesar di dunia. Harga CPO dunia berdampak signifikan terhadap volume ekspor CPO serta pertumbuhan ekonomi Indonesia. Peramalan harga CPO secara akurat sangat penting untuk membantu pengusaha, investor, produsen dan juga pemerintah dalam perencanaan ekonomi dengan mempertimbangkan perubahan dan fluktuasi harga. Penelitian ini menggunakan data harga bulanan CPO dunia dari Januari 2003 hingga April 2023 yang diperoleh dari rata-rata harga harian CPO yang bersumber dari kompilasi data ISTA Mielke GmbH, Oil World, US Department of Agriculture dan World Bank. Metode yang digunakan dalam penelitian ini adalah Autoregressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM) dan Hybrid ARIMA-LSTM. Penggunaan ketiga metode ini didasarkan pada kondisi data yang berbentuk linear dan nonlinear. Metode LSTM terbukti lebih efektif dalam memprediksi harga CPO dunia dibanding metode ARIMA dan Hybrid ARIMA-LSTM. Model LSTM terbaik adalah model dengan hyperparameter epoch 100, batch size 1, dan learning rate 0,001. Model ini menghasilkan nilai Mean Absolute Percentage Error 2,33% dan Root Mean Squared Error 34,708. Hal ini karena model LSTM lebih stabil dalam mempelajari pola data deret waktu.id
dc.description.abstractCrude Palm Oil (CPO) is one of Indonesia's main export commodities in the agricultural sector. Indonesia is also the largest producer of CPO in the world. The world CPO prices have a significant impact on Indonesia's export volume of CPO as well as its economic growth. It is essential to forecast CPO prices accurately to assist entrepreneurs, investors, producers, and the government in economic planning for the future while considering changes and price fluctuations. This study uses world CPO monthly price data from January 2003 to April 2023 which is compilation of data from ISTA Mielke GmbH, Oil World, US Department of Agriculture and World Bank. The methods used in this research are Autoregressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), and Hybrid ARIMA-LSTM. The use of these methods is based on linear and nonlinear data conditions. Due to its ability to learn the data pattern well, the LSTM was found to be more effective in forecasting world CPO prices. The best LSTM model in this study is a model with hyperparameter epoch 100, batch size 1, and learning rate 0,001. This model produces a MAPE and an RMSE value of 2,33% and 34,708.id
dc.language.isoidid
dc.publisherIPB Universityid
dc.titleAplikasi Metode ARIMA, LSTM, dan Hybrid ARIMA-LSTM pada Peramalan Harga Crude Palm Oil (CPO) Duniaid
dc.typeUndergraduate Thesisid
dc.subject.keywordARIMAid
dc.subject.keywordcrude palm oil pricesid
dc.subject.keywordforecastingid
dc.subject.keywordhybrid ARIMA-LSTMid
dc.subject.keywordLSTMid
Appears in Collections:UT - Statistics and Data Sciences

Files in This Item:
File Description SizeFormat 
Cover.pdf
  Restricted Access
Cover1.62 MBAdobe PDFView/Open
G14190091_Elisabeth Kurniasari Marcela Uskono.pdf
  Restricted Access
Fulltext3.26 MBAdobe PDFView/Open
Lampiran.pdf
  Restricted Access
Lampiran1.29 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.