River Water Suspended Sediment Predictive Analytics Using Artificial Neural Network and Convolutional Neural Network Approach: A Review

• Main Authors: Khan Q., Hayder G., Al-Zwainy F.M.S.
• Other Authors: 58309988500
• Format: Conference Paper
• Published: Springer Nature 2024
• Subjects: ANN; CNN; Langat river; MAE; R²; RMSE; Suspended sediment

For water resource management and water quality challenges, estimating suspended sediment is crucial. It demands accurate data and information on suspended sediment concentrations (SSC). Because real sampling can be difficult during severe weather and certain old approaches will not yield enough data, engineers are developing new accurate forecasting technologies. The aim of this study is to see if machine learning techniques like convolutional neural network (CNN) and artificial neural network (ANN) could be utilized to estimate SSC in Malaysia�s Langat stream. The CNN is a form of machine learning method that has not gotten much attention around SSC prediction. The prediction model created in this work is intended for use in the water quality monitoring of Langat stream in Malaysia. River discharge and suspended solids will be the input variables for the models. Both models will be analyzed using the three criteria for performance including root mean square error (RMSE), coefficient of determination (R2), and mean absolute error (MAE) to find which is more accurate in predicting the SSC for this river. The model with the top performance will have the lowest MAE and RMSE values as well as the high value of R2. This study will contribute to demonstrating how machine learning may be used to forecast future suspended sediment concentrations in rivers. � 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.