Advanced Machine Learning Techniques for Predicting Nha Trang Shorelines

Nha Trang Coast is located in the South Central Vietnam and coastal erosion has occurred rapidly recent years. Hence it crucial to accurately monitor shoreline changes for better management reduction of risks communities. In this paper, we explored a statistical forecasting model, Seasonal Auto-regressive Integrated Moving Average (SARIMA), two Machine Learning (ML) models, Neural Network Auto-Regression (NNAR) Long Short-Term Memory (LSTM), predict variations from surveillance camera images. Compared Empirical Orthogonal Function (EOF), most common method used predicting cameras, demonstrate that SARIMA, NNAR LSTM models outperform EOF model significantly terms prediction accuracy. The performance SARIMA comparable both long short-term predictions. results suggest these are highly effective detecting video cameras under extreme weather conditions.