Research on the Development and Application of a Deep Learning Model for Effective Management and Response to Harmful Algal Blooms

Harmful algal blooms (HABs) caused by harmful cyanobacteria adversely impact the water quality in aquatic ecosystems and burden socioecological systems that are based on utilization. Currently, Korea uses Environmental Fluid Dynamics Code-National Institute of Research (EFDC-NIER) model to predict algae conditions respond through HAB alert system. This study aimed establish an additional deep learning effectively blooms. The prediction is a neural network (DNN), which type artificial widely used for prediction. By applying synthetic minority over-sampling technique (SMOTE) resolve imbalance data, DNN showed improved performance during validation predicting number cells. R-squared increased from 0.7 0.78, MAE decreased 0.6, RMSE 0.9 0.7, indicating enhancement model’s performance. Furthermore, regarding levels, 0.18 0.79, 0.2 0.1, 0.3 0.2, as well. According results, constructed data-based reasonably predicted summer when bloom-induced damage occurs accurately levels immediate decision-making. main objective this was develop new technology managing HABs river environments, aiming sustainable future ecosystem.