West texas intermediate crude oil price prediction using arima and lstm models = Prediksi harga minyak mentah west texas intermediate dengan menggunakan model arima dan lstm

Huang, Karin Nathania (2024) West texas intermediate crude oil price prediction using arima and lstm models = Prediksi harga minyak mentah west texas intermediate dengan menggunakan model arima dan lstm. Bachelor thesis, Universitas Pelita Harapan.

	Text (Title) Title.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (219kB)
	Text (Abstract) Abstract.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (769kB)
	Text (TOC) ToC.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (733kB)
	Text (Chapter 1) Chapter1.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (769kB)
	Text (Chapter 2) Chapter2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (998kB)
	Text (Chapter 3) Chapter3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (936kB)
	Text (Chapter 4) Chapter4.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (2MB)
	Text (Chapter 5) Chapter5.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (723kB)
	Text (Bibliography) Bibliography.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (733kB)
	Text (Appendices) Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (4MB)

Abstract

Crude oil is a non-renewable natural resource that is crucial for global economic growth. Changes in oil prices have a substantial impact on various economic aspects, including inflation, commodity prices, and industries dependent on oil consumption. The methods used in this study are Autoregressive Integrated Moving Average (ARIMA), Long Short Term Memory (LSTM), and Autoregressive Integrated Moving Average - Long Short Term Memory (ARIMA- LSTM). The data used in this research is divided into three periods and the entire dataset. The aim of this study is to develop predictive models and compare the ARIMA, LSTM, and ARIMA-LSTM models to find the most accurate model for predicting West Texas Intermediate (WTI) crude oil prices. The results of the study are based on comparisons of MSE, RMSE, and MAPE values. The ARIMA model with the best MSE, RMSE, and MAPE values for the first period is 0.024742, 0.157296, and 0.997816%, for the second period 0.004952, 0.070375, and 2.078522%, for the third period 0.014839, 0.121816, and 1.362598%, and for the entire dataset 0.013679, 0.116960, and 1.658289%. The LSTM model with the best MSE, RMSE, and MAPE values for the first period is 0.877480, 0.936739, and 5.920165%, for the second period 1823.509, 42.70257, and 41.579420%, for the third period 9.278621, 3.046083, and 4.221886%, for the entire dataset in case 1 is 0.013617, 0.116692, and 1.634682%, and in case 2, 0.001512, 0.038888, and 1.635732%. The ARIMA-LSTM model with the best MSE, RMSE, and MAPE values for the first period is 0.469690, 0.6853393, and 5.609322%, for the second period 1795.160, 42.36933, and 41.38892%, for the third period 10.15374, 3.186494, and 4.396291%, for the entire dataset in case 1 is 0.008061, 0.089784, and 1.653329%, and in case 2, 0.047023, 0.216848, and 0.8533330%. In conclusion, the three data periods, the ARIMA model predicts better. However, the overall results of the study indicate that the ARIMA-LSTM model can predict more accurately than the other models. / Minyak mentah adalah sumber daya alam yang tidak dapat diperbarui dan sangat penting bagi pertumbuhan ekonomi global. Perubahan harga minyak memiliki dampak substansial pada berbagai aspek ekonomi, termasuk inflasi, harga barang, dan industri yang bergantung pada konsumsi bahan bakar minyak. Metode yang akan digunakan adalah Autoregressive Integrated Moving Average (ARIMA), Long Short Term Memory (LSTM), dan Autoregressive Integrated Moving Average - Long Short Term Memory (ARIMA-LSTM). Dalam penelitian ini, data yang digunakan adalah data yang telah dibagi menjadi 3 periode data dan seluruh data. Tujuan dari penelitian ini adalah membuat model prediksi serta membandingkan model ARIMA, LSTM, dan ARIMA-LSTM untuk mencari model yang lebih akurat untuk memprediksi harga minyak mentah West Texas Intermediate (WTI). Hasil penelitian didapat berdasarkan perbandingan dari nilai MSE, RMSE dan MAPE. Model ARIMA dengan nilai MSE, RMSE dan MAPE terbaik pada periode ke-1 adalah 0.024742, 0.157296 dan 0.997816%, periode ke-2 0.004952, 0.070375 dan 2.078522%, periode ke-3 0.014839, 0.121816 dan 1.362598%, untuk seluruh data 0.013679, 0.116960 dan 1.658289%. Model LSTM dengan nilai MSE, RMSE dan MAPE terbaik pada periode ke-1 adalah 0.877480, 0.936739 dan 5.920165%, pada periode ke-2 1823.509, 42.70257 dan 41.579420%, pada periode ke-3 9.278621, 3.046083 dan 4.221886%, untuk seluruh data pada kasus 1 adalah 0.013617, 0.116692 dan 1.634682%, dan pada kasus 2, 0.001512, 0.038888 dan 1.635732%. Model ARIMA-LSTM dengan nilai MSE, RMSE dan MAPE terbaik pada periode ke-1 adalah 0.469690, 0.6853393 dan 5.609322%, pada periode ke-2 1795.160, 42.36933 dan 41.38892%, pada periode ke-3 10.15374, 3.186494 dan 4.396291%, untuk seluruh data pada kasus 1 adalah 0.008061, 0.089784 dan 1.653329%, dan pada kasus 2, 0.047023, 0.216848 dan 0.8533330%. Dapat disimpulkan bahwa untuk ketiga periode data model ARIMA memprediksi lebih baik. Namun, hasil penelitian secara keseluruhan menunjukkan bahwa model ARIMA-LSTM dapat memprediksi lebih akurat dari model lainnya.

Actions (login required)

View Item