Comparison of Artificial Intelligence (AI) Based Models for Sediment Transport Prediction Using SWOT and Statistical Analyses

Abstract

The dynamics involved in sediment scour are complicated to create a general empirical optimization algorithm to offer reliable sediment load estimation. The existing study was conducted to analyse the architectures of assorted artificial intelligence (AI) based model to forecast suspended sediment load in fluvial system. An in-depth study on traditional approach including Artificial Neural Network (ANN), Adaptive NeuroFuzzy Inference System (ANFIS), and Genetic Programming (GP) was carried out. The goal of this study is to evaluate the performance of AI-based models from various research using SWOT and statistical analyses. Three statistical measures of model prediction accuracy including coefficient of correlation (R), root mean square error (RMSE), and mean absolute error (MAE) were used. The results revealed that the SVM and ANFIS models outperformed the other soft computing and conventional models. It is concluded that the SVM and ANFIS models are preferred and may be successfully used to estimate the suspended sediment concentration for the research area.