| dc.description.abstract | Dissolved Oxygen (DO) is a essential parameter in the aquascape ecosystem that affects the health of aquatic biota. Changes in dissolved oxygen levels are influenced by various factors such as pH, total dissolved solids (TDS), temperature, and historical dissolved oxygen data. Measuring dissolved oxygen levels using sensors and manual methods has limitations in analyzing long-term trends and anticipating changes proactively. Therefore, this study aims to develop a model to predict dissolved oxygen levels by implementing the Long Short-Term Memory (LSTM) algorithm, which has been proven reliable in handling time-series data. The dataset used in this research was collected from the Harvest aquascape store, where data was recorded over approximately two months with a sampling interval of every 10 minutes. The LSTM model was trained using data that had undergone a pre-processing phase and was evaluated using the Mean Squared Error (MSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Root Mean Squared Error (RMSE), R-Squared (R^2), and Accuracy metrics. The results show that the LSTM model can predict DO levels with a high degree of accuracy. The evaluation results show that the model achieved an MSE of 0.4929, MAE of 0.4647, MAPE of 0.0239, RMSE of 0.7021, R^2 of 0.9690, and an accuracy of 97.61%. This research demonstrates that the model can serve as a useful tool for aquascape practitioners in monitoring and managing DO levels more effectively, thereby maintaining the balance of the aquascape ecosystem. | en_US |
