Nosius Luaran¹, Rayner Alfred¹*, Xu Fengchang² and Haviluddin³

¹Creative Advanced Machine Intelligence Research Centre, Faculty of Computing and Informatics, Universiti Malaysia Sabah, Jalan UMS, 88400 Kota Kinabalu, Sabah, Malaysia
²Department of Information Engineering, Shandong Light Industry Vocational College, Zibo City, Shandong Province, China
³Department of Informatics, Faculty of Engineering, Universitas Mulawarman, East Kalimantan, Indonesia

*Corresponding Author: Rayner Alfred, Creative Advanced Machine Intelligence Research Centre, Faculty of Computing and Informatics, Universiti Malaysia Sabah, Jalan UMS, 88400 Kota Kinabalu, Sabah, Malaysia.

Received: June 13, 2025; Published: July 10, 2025

Abstract

Generating accurate and reliable tidal forecasts is very crucial to support routine coastal decision-making. The forecasting of univariate time series behavior becomes more challenging and requires state-of-the-art techniques to achieve realistic forecasting outcome. However, the larger the initial hyperparameters space that must be searched, the more iterations the automated hyperparameters tuning technique should have. As a result of the complexity to tune the deep learning architecture in producing an optimal and consistent result, tuning several hyperparameters to improve deep learning results in learning time series data is highly needed. Therefore, it is necessary to investigate the best approach to fine tune machine learning algorithms that have multiple hyperparameters (e.g., deep Long Short-Term Memory (LSTM)) to learn and gather collective knowledge about time series sea water tidal shift data that could be used to make better forecasts for the individual time series. Thus, the aim of this paper is to investigate the effects of varying the values of the hyperparameters of the LSTM architecture on the forecasting accuracy of the sea water tidal height variation under nonextreme conditions. In this paper, we empirically evaluate the proposed LSTM forecasting framework to model the tidal dataset based on the RMSE measurements as the hyperparameters change. These hyperparameters are divided into static hyperparameters (e.g., mini batch size, number of epochs, number of iteration and percentage of neuron dropout) and dynamic hyperparameters (e.g., number of layers, number of hidden units, learning rate and L2 regularization). The initial static hyperparameters are used to estimate the optimal values of the dynamic parameters using the Bayesian Optimization method. Based on the results, we highlighted some findings that will aid in improving the performance of the LSTM model for tidal forecasting for hourly short term sea water level and these allow users to choose the best static and dynamic hyper-parameters combination that can produce the optimum outcomes. Finally, the paper is concluded by suggesting some of the extended works that can be performed to improve the results of forecasting the sea water tidal shift.

Keywords: Forecasting Sea Water Tidal Shift; Long Short-Term Memory; Time-Series Data; Hyperparameters Setting; Deep Learning; Bayesian Optimization

References

1. Yaqoob I., et al. “Internet of things architecture: Recent advances, taxonomy, requirements, and open challenges”. IEEE Wireless Communications 3 (2017): 10-16.
2. Canizo M., et al. “Multi-head cnn-rnn for multi-time series anomaly detection: An industrial case study”. Neurocomputing 363 (2019): 246-260.
3. Wan R., et al. “Multivariate temporal convolutional network: A deep neural networks approach for multivariate time series forecasting”. Electronics8 (2019).
4. Abubakar AG., et al. “A review of modelling approaches on tidal analysis and prediction”. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLII-4/W16 (2019): 23-34.
5. Van Houdt G., et al. “N´apoles: A review on the long short-term memory model”. Artificial Intelligence Review 53 (2020): 5929-5955.
6. Yamak PT., et al. “A comparison between arima, lstm, and gru for time series forecasting”. In: Proceedings of the 2019 2nd International Conference on Algorithms, Computing and Artificial Intelligence. ACAI (2019): 49-55. Association for Computing Machinery, New York, NY, USA (2019).
7. Peter G and Matskevichus M. “Hyperparameters tuning for machine learning models for time series forecasting”. In: 2019 Sixth International Conference on Social Networks Analysis, Management and Security (SNAMS) (2019): 328-332.
8. Hewamalage H., et al. “Recurrent neural networks for time series forecasting: Current status and future directions”. International Journal of Forecasting1 (2021): 388-427.
9. Hewamalage H., et al. “Recurrent neural networks for time series forecasting: Current status and future directions”. International Journal of Forecasting1 (2021): 388-427.
10. Dixon M and London J. “Financial forecasting with a-rnns: A time series modeling approach”. Frontiers in Applied Mathematics and Statistics 6 (2021): 59.
11. Casado-Vara R., et al. “Web traffic time series forecasting using lstm neural networks with distributed asynchronous training”. Mathematics 4 (2021).
12. Le XH., et al. “Application of long short-term memory (lstm) neural network for flood forecasting”. Water 7 (2019).
13. Zhang Q., et al. “Prediction of sea surface temperature using long short-term memory”. IEEE Geoscience and Remote Sensing Letters10 (2017): 1745-1749.
14. Van Houdt G., et al. “El ni˜nosouthern oscillation forecasting using complex networks analysis of lstm neural networks”. Artificial Life and Robotics 24 (2019): 445-451.
15. , et al. “A forecasting model for wave heights based on a long short-term memory neural network”. Acta Oceanologica Sinica 40 (2021): 62-69.
16. Kelany O., et al. “Deep learning model for financial time series prediction”. In: 2020 14th International Conference on Innovations in Information Technology (IIT) (2020): 120-125.
17. Zhou S., et al. “An optimized heterogeneous structure lstm network for electricity price forecasting”. IEEE Access 7 (2019): 108161-108173.
18. , et al. “Forecasting tourist arrivals via random forest and long short-term memory”. Cognitive Computation 13 (2021): 125-138.
19. Saini U., et al. “Univariant time series forecasting of agriculture load by using lstm and gru rnns”. In: 2020 IEEE Students Conference on Engineering Systems (SCES) (2020): 1-6.
20. Wu CH., et al. “A new forecasting framework for bitcoin price with lstm”. In: 2018 IEEE International Conference on Data Mining Workshops (ICDMW) (2018): 168-175.
21. Zhou Y., et al. “Lstm based network flow forecasting in high-speed railway data network”. In: 2019 IEEE International Conferences on Ubiquitous Computing Communications (IUCC) and Data Science and Computational Intelligence (DSCI) and Smart Computing, Networking and Services (SmartCNS) (2019): 71-75.
22. Sarah A., et al. “Lstm model to forecast time series for ec2 cloud price”. In: 2018 IEEE 16th Intl Conf on Dependable, Autonomic and Secure Computing, 16th Intl Conf on Pervasive Intelligence and Computing, 4th Intl Conf on Big Data Intelligence and Computing and Cyber Science and Technology Congress (2018): 1085-1088.
23. Rauf HT., et al. “Time series forecasting of covid-19 transmission in asia pacific countries using deep neural networks”. Personal and Ubiquitous Computing (2021): 1-18.
24. , et al. “Forecasting hotel reservations with long short-term memory-based recurrent neural networks”. International Journal of Data Science and Analytics 9 (2020): 77-94.
25. Pala Z and Atici R. “Forecasting sunspot time series using deep learning methods”. Solar Physics 294 (2019): 50.
26. Yang H., et al. “A network traffic forecasting method based on sa optimized arima-bp neural network”. Computer Networks 193 (2021): 108102.
27. Helli SS., et al. “Short-term forecasting covid-19 cases in turkey using long short-term memory network”. 2020 Medical Technologies Congress (TIPTEKNO) 1-4 (2020).
28. Karsanti HT., et al. “Deep learning-based patient visits forecasting using long short term memory. In: 2019 International Conference of Artificial Intelligence and Information Technology (ICAIIT) (2019): 344-349.
29. Yao S., et al. “High-frequency stock trend forecast using lstm model”. In: 2018 13th International Conference on Computer Science Education (ICCSE) (2018): 1-4.
30. Han Y. “A forecasting method of pharmaceutical sales based on arima-lstm model”. In: 2020 5th International Conference on Information Science, Computer Technology and Transportation (ISCTT) (2020): 336-339.
31. Yadav A., et al. “Optimizing lstm for time series prediction in Indian stock market”. Procedia Computer Science 167 (2020): 2091-2100.
32. Chicco D., et al. “The coefficient of determination r-squared is more informative than smape, mae, mape, mse and rmse in regression analysis evaluation”. PeerJ Computer Science 7 (2021).
33. Alfred R and Obit JH. “The roles of machine learning methods in limiting the spread of deadly diseases: A systematic review”. Heliyon 7 (2021).

Citation

Citation: Rayner Alfred., et al. “Assessment of the Optimization of Hyperparameters in Deep LSTM for Time Series Sea Water Tidal Shift".Acta Scientific Computer Sciences 7.3 (2025): 03-15.

Copyright

Copyright: © 2025 Rayner Alfred., et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.