PREDIKSI HARGA EMAS BERDASARKAN INFLASI DAN SUKU BUNGA MENGGUNAKAN ALGORITMA XGBOOST DAN SUPPORT VECTOR REGRESSION (SVR) DENGAN BAYESIAN OPTIMIZATION

Alfa, Aniysah Fauziyyah (2025) PREDIKSI HARGA EMAS BERDASARKAN INFLASI DAN SUKU BUNGA MENGGUNAKAN ALGORITMA XGBOOST DAN SUPPORT VECTOR REGRESSION (SVR) DENGAN BAYESIAN OPTIMIZATION. Undergraduate thesis, Universitas Pembangunan Nasional Veteran Jawa Timur.

	Text (Cover) 21083010083_Cover.pdf Download (1MB)
	Text (Bab 1) 21083010083_Bab 1.pdf Download (274kB)
	Text (Bab 2) 21083010083_Bab 2.pdf Restricted to Registered users only until 19 June 2028. Download (598kB) \| Request a copy
	Text (Bab 3) 21083010083_Bab 3.pdf Restricted to Registered users only until 19 June 2028. Download (387kB) \| Request a copy
	Text (Bab 4) 21083010083_Bab 4.pdf Restricted to Registered users only until 19 June 2028. Download (5MB) \| Request a copy
	Text (Bab 5) 21083010083_Bab 5.pdf Download (222kB)
	Text (Daftar Pustaka) 21083010083_Daftar Pustaka.pdf Download (239kB)
	Text (Lampiran) 21083010083_Lampiran.pdf Restricted to Registered users only until 19 June 2028. Download (294kB) \| Request a copy

Abstract

Gold prices experience significant fluctuations, especially in Indonesia, influenced by various factors such as inflation rates and interest rates. Therefore, gold price prediction is very necessary in determining financial strategies. This study develops a gold price prediction model by comparing the performance of the XGBoost and Support Vector Regression (SVR) algorithms, and using Bayesian Optimization to determine the best combination of parameters. The research process includes data pre-processing, feature engineering with the addition of technical indicators, and implementation of recursive predictions for the next 30 days. Based on the evaluation using MAE, RMSE, and MAPE metrics, SVR proved to be superior to XGBoost in predicting gold prices, with MAE results of IDR 2,762.94, RMSE of IDR 3,419.14, and MAPE of 0.19%. In contrast, XGBoost showed much lower performance with MAE reaching IDR 106,408.79, RMSE of IDR 121,438.00, and MAPE of 7.13%. This shows that SVR is more accurate and precise for short-term predictions. Bayesian Optimization is proven to be effective in improving the accuracy of SVR by finding the optimal parameter combination (C = 49.999; epsilon = 0.01; gamma = 0.0071). Technical features have a significant impact on the accuracy of short- to medium-term predictions, while macroeconomic factors remain relevant for long-term analysis. The results of this study can provide benefits for investors and policymakers in developing more adaptive and precise gold investment strategies amid market dynamics.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Prasetya, Dwi Arman	NIDN0005128001	arman.prasetya.sada@upnjatim.ac.id
Thesis advisor	Sugiarto, Sugiarto	NIDN0714028703	sugiarto.if@upnjatim.ac.id

Subjects:

H Social Sciences > HA Statistics
Q Science > QA Mathematics
Q Science > QA Mathematics > QA76.6 Computer Programming

Divisions:

Faculty of Computer Science > Departemen of Data Science

Depositing User:

Aniysah Fauziyyah Alfa

Date Deposited:

20 Jun 2025 02:22

Last Modified:

20 Jun 2025 02:22

URI:

https://repository.upnjatim.ac.id/id/eprint/38511