Research Article

Prediction of gold price fluctuations with artificial neural networks: a comprehensive approach

Volume: 14 Number: 1 June 30, 2026
EN

Prediction of gold price fluctuations with artificial neural networks: a comprehensive approach

Abstract

Gold is an important investment instrument of choice around the world. At the same time, its value and growth potential are of great interest. The appeal of this precious metal lies in its intrinsic value and potential for lucrative growth over time, making it a subject of intense interest and analysis among investors and economists alike. One of the key determinants of economic and social dynamics is the volatility of exchange rates and the gold market. Therefore, understanding and correctly interpreting the fluctuations of exchange rates and the gold market is of great importance for both individual investors and corporate firms. In this study, the factors affecting gold prices were investigated and the correlation between the values found and gold prices was revealed, and it was aimed to predict the change in gold price in the future with Artificial Neural Networks (ANN). The data set used in the study consists of data between January 2, 2020, and March 15, 2024. Since the markets were closed on holidays and weekends, these dates were separated and a total of 1005 days of data were used. The developed model was created and tested with different layers and neurons using MATLAB Neural Net Fitting. Results were evaluated using Coefficient of Determination (R²), Mean Squared Error (MSE) and Mean Absolute Error (MAE). The best success rate was obtained by using Levenberg-Marquardt algorithm, hyperbolic tangent sigmoid activation function in hidden layers and “purelin” linear activation function in the output layer, 4 layers and 3 neurons and 94.06% success rate was observed.

Keywords

artificial neural networks, gold prices, data analysis

Ethical Statement

The authors state that this research adheres to the ethical standards. This research does not involve either human participants or animals.

Thanks

The authors state that this research adheres to the ethical standards. This research does not involve either human participants or animals.

References

  1. Topçu, A., Altın Fiyatlarını Etkileyen Faktörler. 2010, SPK Araştırma Raporu.
  2. Söylemez, Y., Prediction of Gold Prices Using Multilayer Artificial Neural Networks Method, Sosyoekonomi, 28, 46, 271-291, (2020).
  3. Arz Talep Bilgisi. 2024; Available from: https://www.gold.org/.
  4. Öndeş, H. and A. Oğuzlar, Estimated Gold Price (TL/Kg) With Artificial Neural Networks, Akademik Bakış Dergisi, 249-262, (2019).
  5. Kaya, E. Training Of Neural Network By Using ABC Algorithm, PSO And FPA For Prediction Of Gold Price, International Black Sea Modern Scientific Research Congress, Rize, (2022).
  6. Keskin, M. and A. Yücel, BIST 100 Endeksi İle Altın Fiyatları İlişkisinin Yapay Sinir Ağları Yöntemiyle Belirlenmesi (1988-2020), MANAS Journal of Social Studies, 11, 2, (2022).
  7. Yüksel, R. and S. Akkoç, Forecasting Gold Prices By Using Artificial Neural Network And An Application, Doğuş Üniversitesi Dergisi, 17, 1, 39-50, (2016).
  8. Harman, G., Using Different Neural Networks Learning Algorithms Next Rate Prediction of Gold/Dollar Parity, Veri Bilimi Dergisi, (2022).
  9. Kocatepe, C.İ. and O. Yıldız, Forecasting Of The Direction Changes In The Gold Price In Turkey With Artificial Neural Network By Using Economic Indices, Düzce Üniversitesi Bilim ve Teknoloji Dergisi, (2016).
  10. Artificial Neural Network - Basic Concepts. 2024; Available from: https://www.tutorialspoint.com/artificial_neural_network/artificial_neural_network_basic_concepts.htm.
APA
Eldem, A., & Ay Gülşen, G. (2026). Prediction of gold price fluctuations with artificial neural networks: a comprehensive approach. MANAS Journal of Engineering, 14(1), 21-29. https://doi.org/10.51354/mjen.1742267
AMA
1.Eldem A, Ay Gülşen G. Prediction of gold price fluctuations with artificial neural networks: a comprehensive approach. MJEN. 2026;14(1):21-29. doi:10.51354/mjen.1742267
Chicago
Eldem, Ayşe, and Gülistan Ay Gülşen. 2026. “Prediction of Gold Price Fluctuations With Artificial Neural Networks: A Comprehensive Approach”. MANAS Journal of Engineering 14 (1): 21-29. https://doi.org/10.51354/mjen.1742267.
EndNote
Eldem A, Ay Gülşen G (June 1, 2026) Prediction of gold price fluctuations with artificial neural networks: a comprehensive approach. MANAS Journal of Engineering 14 1 21–29.
IEEE
[1]A. Eldem and G. Ay Gülşen, “Prediction of gold price fluctuations with artificial neural networks: a comprehensive approach”, MJEN, vol. 14, no. 1, pp. 21–29, June 2026, doi: 10.51354/mjen.1742267.
ISNAD
Eldem, Ayşe - Ay Gülşen, Gülistan. “Prediction of Gold Price Fluctuations With Artificial Neural Networks: A Comprehensive Approach”. MANAS Journal of Engineering 14/1 (June 1, 2026): 21-29. https://doi.org/10.51354/mjen.1742267.
JAMA
1.Eldem A, Ay Gülşen G. Prediction of gold price fluctuations with artificial neural networks: a comprehensive approach. MJEN. 2026;14:21–29.
MLA
Eldem, Ayşe, and Gülistan Ay Gülşen. “Prediction of Gold Price Fluctuations With Artificial Neural Networks: A Comprehensive Approach”. MANAS Journal of Engineering, vol. 14, no. 1, June 2026, pp. 21-29, doi:10.51354/mjen.1742267.
Vancouver
1.Ayşe Eldem, Gülistan Ay Gülşen. Prediction of gold price fluctuations with artificial neural networks: a comprehensive approach. MJEN. 2026 Jun. 1;14(1):21-9. doi:10.51354/mjen.1742267