IPB University Logo

SCIENTIFIC REPOSITORY

IPB University Scientific Repository collects, disseminates, and provides persistent and reliable access to the research and scholarship of faculty, staff, and students at IPB University

AI Repository
 
Building and Categories


      View Item 
      •   IPB Repository
      • Final Assignments
      • Undergraduate Final Assignments
      • UF - School of Data Science, Mathematic and Informatics
      • UF - Mathematics
      • View Item
      •   IPB Repository
      • Final Assignments
      • Undergraduate Final Assignments
      • UF - School of Data Science, Mathematic and Informatics
      • UF - Mathematics
      • View Item
      JavaScript is disabled for your browser. Some features of this site may not work without it.

      Peramalan Harga Emas (XAU/USD): Perbandingan Akurasi Model ARIMA dan Hybrid ARIMA-LSTM

      Thumbnail
      View/Open
      Cover (1.861Mb)
      Fulltext (3.600Mb)
      Lampiran (5.519Mb)
      Date
      2026
      Jenis/Type
      Skripsi
      Subtype
      Undergraduate Theses
      Author
      Sari, Titing Sri Angella
      Ardana, Ngakan Komang Kutha
      Budiarti, Retno
      Metadata
      Show full item record
      Abstract
      Emas merupakan salah satu komoditas penting dalam perekonomian global yang berfungsi sebagai instrumen investasi sekaligus alat lindung nilai terhadap inflasi dan fluktuasi nilai tukar. Namun, dalam beberapa dekade terakhir, harga emas menunjukkan tingkat volatilitas yang tinggi sehingga pergerakannya sulit diprediksi. Penelitian ini bertujuan untuk menentukan model terbaik antara ARIMA dan hybrid ARIMA-LSTM dalam meramalkan harga emas (XAU/USD) menggunakan data harian periode Januari 2004 hingga April 2024. Model ARIMA yang digunakan adalah ARIMA (5,1,6) dengan drift untuk menangkap pola linier, sedangkan LSTM digunakan untuk memodelkan pola nonlinier pada residu ARIMA. Evaluasi kinerja model dilakukan menggunakan RMSE, MAE, dan MAPE pada data uji. Hasil penelitian menunjukkan bahwa model hybrid dengan Min-Max Scaler hanya memberikan peningkatan akurasi yang terbatas dengan nilai MAPE sebesar 5.9604%, sedangkan model hybrid dengan Robust Scaler menggunakan quantile range = (5,95) menunjukkan kinerja yang lebih baik dengan nilai MAPE sebesar 5.6750%. Model hybrid dengan Robust Scaler kemudian digunakan untuk peramalan 10 hari ke depan dan menghasilkan nilai MAPE sebesar 0.84%, yang menunjukkan tingkat akurasi sangat baik. Dengan demikian, model hybrid ARIMA-LSTM dengan Robust Scaler merupakan model terbaik karena mampu menghasilkan peramalan yang lebih akurat serta lebih efektif dalam menangkap pola nonlinier pada data harga emas.
       
      Gold is one of the most important commodities in the global economy, serving both as an investment instrument and a hedge against inflation and exchange rate fluctuations. However, over the past decades, gold prices have exhibited high volatility, making their movements difficult to predict. This study aims to determine the best forecasting model between ARIMA and hybrid ARIMA-LSTM for predicting gold prices (XAU/USD) using daily data from January 2004 to April 2024. The ARIMA (5,1,6) model with drift was employed to capture linear patterns, while LSTM was used to model nonlinear patterns in the ARIMA residuals. Model performance was evaluated using RMSE, MAE, and MAPE on the testing data. The results showed that the hybrid model with Min-Max Scaler provided only limited improvement in forecasting accuracy, yielding a MAPE value of 5.9604%, whereas the hybrid model with Robust Scaler using a quantile range of (5,95) achieved better performance with a MAPE value of 5.6750%. The hybrid model with Robust Scaler was subsequently used to forecast gold prices for the next 10 days and produced a MAPE value of 0.84%, indicating very high forecasting accuracy. Therefore, the hybrid ARIMA-LSTM model with Robust Scaler was identified as the best model, as it generated more accurate forecasts and was more effective in capturing nonlinear patterns in gold price data.
       
      URI
      http://repository.ipb.ac.id/handle/123456789/176585
      Collections
      • UF - Mathematics [164]

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository
        

       

      Browse

      All of IPB RepositoryCollectionsBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

      My Account

      Login

      Application

      google store

      Copyright © 2020 Library of IPB University
      All rights reserved
      Contact Us | Send Feedback
      Indonesia DSpace Group 
      IPB University Scientific Repository
      UIN Syarif Hidayatullah Institutional Repository
      Universitas Jember Digital Repository