Archives
 
Year 2025
 
Vol 66, Issue 3, [06 - 2025]

Predicting elevation values using Gated Recurrent Unit (GRU) neural networks

DOI:10.46326/JMES.2025.66(3).03

  • Affiliations:

    1 Faculty of Geomatics and Land Administration, Hanoi University of Mining and Geology, Hanoi, Vietnam
    2 Geodesy and Environment Research Group, Hanoi University of Mining and Geology, Hanoi, Vietnam

  • *Corresponding:
    This email address is being protected from spambots. You need JavaScript enabled to view it.
  • Received: 8th-Jan-2025
  • Revised: 24th-July-2025
  • Accepted: 15th-May-2025
  • Online: 1st-June-2025
Pages: 29 - 38
Views: 48
View onlineView online
Downloads: 3
Downdload PDF
★★★★★Rating: , Total rating: 0
  • ★
  • ★
  • ★
  • ★
  • ★
Yours rating

Abstract:

This paper focuses on the development and evaluation of a GNSS height point prediction model based on the Gated Recurrent Unit (GRU), an advanced variant of recurrent neural networks designed for time series data. The input data consists of height coordinate sequences (h) extracted from GNSS data processed at the CPHU station using the specialized Gamit/Globk software, achieving millimeter-level accuracy. Unlike many previous studies, the GRU model in this research was trained on the entire continuous data sequence without splitting it into separate training and testing sets. The prediction performance was assessed over the period from March 9th, 2022, to March 17th, 2022. The model was tested using two optimization algorithms, Adam and SGD, in combination with two commonly used loss functions, MSE and Huber. The performance of each configuration was evaluated using metrics including MSE, RMSE, MAE, and the coefficient of determination (R²). The results indicate that the GRU model combined with the Adam optimizer and MSE loss function yielded the best prediction performance, with an R² value of approximately 0.54 and an MAE of 4.24 mm. In contrast, using SGD and Huber led to a significant decrease in performance, with R² values ranging from only 0.33÷0.42. Additionally, the size of the sliding window (lag) also influenced the model's predictive capability: a smaller window size (lag = 10) allowed the model to better adapt to noisy data.

How to Cite
Nguyen, N.Viet and ., T.Gia Nguyen 2025. Predicting elevation values using Gated Recurrent Unit (GRU) neural networks (in Vietnamese). Journal of Mining and Earth Sciences. 66, 3 (Jun, 2025), 29-38. DOI:https://doi.org/10.46326/JMES.2025.66(3).03.
References

Ceder, N. (2018). The quick Python book: Simon and Schuster.

Chen, D. Y. (2017). Pandas for everyone: Python data analysis: Addison-Wesley Professional.

Cho, J. M. (2015). Estimation of the crustal deformation caused by earthquake and its use in updating published coordinates of geodetic control points-A case study of the 2011 Tohoku Earthquake’s impact in South Korea. Journal of the Korean Society of Surveying, Geodesy, Photogrammetry, Cartography, 33(6), 485-495.

Chollet, F. (2021). Deep learning with Python: Simon and Schuster.

Construction, M. o. (2019). Trouble underground - Land Subsidence in the Mekong Delta. In (pp. https://www.eda.admin.ch/dam/countries/countries-content/vietnam/vn/sdc-publications/ground-subsidence_VN.pdf): Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH.

Crocetti, L., Schartner, M., Schneider, R., Schindler, K. and Soja, B. (2024). Modelling and Analysing GNSS Displacements with Machine Learning and Environmental Variables. JGR: Solid Earth. https://doi.org/10.22541/essoar.172838522.21948627/v1.

Goodfellow, I., Bengio, Y. and Courville, A. (2016). Regularization for deep learning. Deep learning, 216-261.

Gupta, P. and Bagchi, A. (2024). Introduction to NumPy. In Essentials of Python for Artificial Intelligence and Machine Learning (pp. 127-159): Springer.

Inazu, D., Waseda, T., Hibiya, T. and Ohta, Y. (2016). Assessment of GNSS-based height data of multiple ships for measuring and forecasting great tsunamis. Geoscience Letters, 3, 1-13.

Karim, M. R. (2018). TensorFlow: Powerful Predictive Analytics with TensorFlow: Predict valuable insights of your data with TensorFlow: Packt Publishing Ltd.

Khánh, T. and Dương, P. Đ. (2011). Phân tích độ ổn định hệ thống mốc độ cao cơ sở trong quan trắc lún công trình. Tạp chí Khoa học Đo đạc và Bản đồ, 8(8), 52-56.

Kiani, M. (2020). A precise machine learning aided algorithm for land subsidence or upheave prediction from GNSS time series. arXiv preprint arXiv:.03772.

Kingma, D. P. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.

Le D. T., Quoc, H. N. D., and Trong, N. G. (2024). Exploring the training results of machine learning models using different batch sizes and epochs: A case study with GNSS time series data. J. Hydro-Meteorol, 19, 90-99.

Long, N. Q. (2016). Đánh giá khả năng ứng dụng mạng nơ-ron nhân tạo dự báo lún bề mặt mỏ do khai thác hầm lò. Tạp chí Khoa học Kỹ thuật Mỏ-Địa chất, 55, 79-88.

Quốc, L. B. (2022). Dự đoán lún bề mặt do tải trọng đặc trưng khi thi công công trình ngầm trong đất yếu ở TP Hồ Chí Minh. Tạp chí Xây dựng, 90 - 93.

Solak, H. İ. (2024). Prediction of GNSS Velocity Accuracies Using Machine Learning Algorithms for Active Fault Slip Rate Determination and Earthquake Hazard Assessment. Applied Sciences, 15(1), 113.

Tosi, S. (2009). Matplotlib for Python developers: Packt Publishing Ltd.

Trọng, N. G., Nghĩa, N. V., Khải, P. C., Thành, N. H., Hà, L. L., Dũng, V. T., . . . Quang, P. N. (2022). Xác định chuyển dịch trên phạm vi lãnh thổ Việt Nam dựa vào dữ liệu của các trạm CORS thuộc mạng lưới VNGEONET. Tạp chí Khí tượng Thủy văn, 739, 59-66.

Trọng, N. G. and Quý, B. N. (2024). Phân tích chuỗi dữ liệu nghiệm triều sử dụng mạng nơ ron hồi tiếp với nút có cổng (GRU). Tạp chí Khí tượng Thủy văn, 765, 39-46.

Vân Phong, D., Trọng, N. G., Chiến, N. V., Thành, N. H., Hà, L. L., Quân, N. V. and Quang, P. N. (2023). Phân tích chuyển dịch thẳng đứng vỏ Trái đất sử dụng hàm ANN từ kết quả xử lý chuỗi dữ liệu GNSS theo thời gian. Tạp chí Khí tượng Thủy văn, 752, 41-50.  

Other articles