Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/161557| Title: | Prediksi Harga Emas Berjangka Menggunakan Model Visibility Graph |
| Other Titles: | Gold Futures Price Prediction Using the Visibility Graph Model |
| Authors: | Septyanto, Fendy Mangku, I Wayan Suryaputra, Akbar Cautsar |
| Issue Date: | 2025 |
| Publisher: | IPB University |
| Abstract: | Penelitian ini menggunakan model visibility graph untuk memprediksi harga penutupan emas berjangka periode 1 September sampai 14 Oktober. Penelitian ini mencakup model visibility graph berbasis kemiripan dengan model random walk, perhitungan nilai prediksi dengan interpolasi linear, serta perhitungan kelayakan hasil prediksi dengan metode mean absolute percentage error (MAPE). Metode random walk dalam visibility graph digunakan untuk mengetahui titik data yang memiliki nilai kemiripan terbesar terhadap titik data terakhir, yang kemudian kedua titik data tersebut digunakan untuk memprediksi nilai data berikutnya dengan interpolasi linear. Hasil penelitian menunjukkan bahwa metode visibility graph dapat memodelkan harga emas berjangka dan memiliki hasil prediksi yang layak dengan nilai MAPE 0.795%. Hasil yang didapat pada penelitian menunjukkan bahwa prediksi dengan interpolasi linear hanya efektif untuk memprediksi nilai satu data berikutnya, sehingga diperlukan pergeseran data training untuk memprediksi nilai pada waktu-waktu berikutnya.
Kata kunci: emas berjangka, interpolasi linear, prediksi, random walk, visibility graph This study uses a visibility graph model to predict the closing price of gold futures for the period September 1 to October 14. This research includes a similarity-based visibility graph model with a random walk model, calculation of predicted values with linear interpolation, and calculation of the feasibility of prediction results with the mean absolute percentage error (MAPE) method. The random walk method in the visibility graph is used to find out the data point that has the greatest similarity value to the last data point, which then the two data points are used to predict the next data value by linear interpolation. The results showed that the visibility graph method can model gold futures prices and has decent prediction results with a MAPE value of 0.795%. The results obtained in the study show that prediction with linear interpolation is only effective for predicting the value of the next one data, so a shift in training data is needed to predict the value at the next time. Keywords: gold futures, linear interpolation, prediction, random walk, visibility graph |
| URI: | http://repository.ipb.ac.id/handle/123456789/161557 |
| Appears in Collections: | UT - Actuaria |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| cover_G5402201023_ebe418d69de942a59dcfe79ecc22ac9a.pdf | Cover | 659.89 kB | Adobe PDF | View/Open |
| fulltext_G5402201023_79f1730b17d24328addb4255d899c19e.pdf Restricted Access | Fulltext | 1.1 MB | Adobe PDF | View/Open |
| lampiran_G5402201023_92092de469cf45ed88f7616432490134.pdf Restricted Access | Lampiran | 1.83 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.