Performance Evaluation of Artificial Intelligence and Heuristic Regression Methods for Rainfall-Runoff Modelling: An Application in Aksu Stream

In this study, taking into account the Aksu Stream data, daily total precipitation (P) and mean flow (Q) values were using time lagged, 8 different Rainfall-Runoff models created runoff value estimated for future. The have been tried with methods performances compared process. Artificial Intelligence (AI) methods, Neural Networks (ANN), Adaptive Neuro Fuzzy Inference System (ANFIS) Heuristic Regression (HR) Multivariate Splines (MARS) Support Vector Machine (SVM) used describing modelling. performance of is determined Root Mean Square Error (RMSE), Correlation Coefficient (R) Absolute (MAE) coefficients. Although AI was very close, lowest error obtained in model ANFIS method (RMSE=3.682, R=0.934, MAE=1.103). HR highest observed on MARS (RMSE=3,101, R=0,952, MAE=1,302). evaluation, it seen that higher than modelling