Prediction of harmful algae blooms in Sabah using deep learning model.

Harmful Algae Blooms (HABs) have emerged as a critical environmental concern, causing significant damage to marine ecosystems and coastal communities worldwide. Sabah, a coastal region in Malaysia, has experienced increased frequency and intensity of HAB events, leading to severe ecological and economic consequences. Traditional methods for detecting and predicting HABs are limited by their manual and time-consuming nature, hindering timely responses and effective mitigation strategies. To address this issue, this study presents an approach for predicting HABs in Sabah using a Deep Learning model. The proposed deep learning model harnesses the potential of artificial intelligence by integrating historical HAB occurrence records. In the first phase, the researcher collected and organized data from the Sabah Fisheries Department, then carefully cleaned and transformed it into a more practical format. After reviewing existing research, the study chose two promising techniques: the Artificial Neural Networks (ANNs) algorithm and the Long Short-Term Memory (LSTM). These techniques were experimented with to determine the most effective one for predicting HABs. Phase three involved rigorous testing of the prediction model using existing data, providing insights into how well it functions with the dataset and revealing the results obtained. In Phase four, the focus shifted to evaluating the model's performance, specifically identifying any reduced accuracy or shortcomings in the prediction percentages. This study adopts various evaluation metrics to assess the model's performance, such as accuracy, precision, recall, and F1-score, on a dataset with distinct training, validation, and testing sets. The authors also conducted usability testing to gather users' feedback on the proposed system. By providing timely predictions, the proposed model empowers stakeholders to take proactive measures to mitigate the ecological and socio-economic impacts of HABs on the marine ecosystem and coastal communities in Sabah. The findings will contribute substantially to the community by improving early warning systems and optimizing resource allocation for monitoring and controlling HABs. [ABSTRACT FROM AUTHOR]