Estimating Tidal Sea levels along the Central Coast of the Western Arabian Gulf using Machine Learning Algorithms (MLAs)

Abstract

Precise tidal forecasting is an academic exercise and a crucial tool for designing and constructing coastal and marine infrastructure. Machine learning algorithms (MLAs) like Random Forest Regression (RF), K-Nearest Neighbors Regression (KN), Gradient Boosting Machines (GBM), and artificial neural networks (ANNs) are powerful data-driven techniques that can be harnessed for this practical purpose. This study utilizes four machine learning algorithms (MLAs), namely (RF), (KN), (GBM), and the Artificial Neural Network - Multilayer Perceptron (ANN-MLP) model, to accurately estimate the tidal levels along the central coast of the western Arabian Gulf, with direct implications for real-world infrastructure planning and construction. Several metrics, such as mean absolute error (MAE), mean squared error (MSE), normalized mean square error (NMSE), mean absolute percentage error (MAPE), correlation coefficient (R), and root mean square error (RMSE), are used to compare how well the MLAs forecast daily tidal levels. The results confirmed the ANN-MLP model's superiority over the other approaches. The ANN-MLP model, a specific type of artificial neural network, yields enhancements in (RMSE) of 8.945% and 19.05%, 14.18% compared to (RF), (KN), and (GBM), respectively, throughout the testing process. The ANN-MLP, being a powerful and versatile machine learning algorithm, demonstrated the best level of accuracy, together with the lowest values for (RMSE). This experiment unequivocally proves that the ANN-MLP method can be utilized as a supervised machine-learning method for accurately forecasting seawater levels of tidal.