Application of Bayesian Neural Networks, Support Vector Machines and Gene Expression Programming Analysis of Rainfall - Runoff Monthly (Case Study: Kakarza River)

Ghorbani, Mohammad Ali; Dehghani, Reza

doi:10.22055/jise.2016.12117

Application of Bayesian Neural Networks, Support Vector Machines and Gene Expression Programming Analysis of Rainfall - Runoff Monthly (Case Study: Kakarza River)

Document Type : Research Paper

Authors

Mohammad Ali Ghorbani

Reza Dehghani

10.22055/jise.2016.12117

Abstract

Simulation of rainfall - runoff process is one of the most important tasks in water resources management and flood control studies. In this study, the rainfall – runoff process over Kakarza river located at Lorestan province, was simulated using the Bayesian neural network and the results were compared with the gene expression and support vector machine models. In this case, different combinations of monthly rainfall and runoff data in period of 1969-2013 were considered as the input data of the models. Four performance criteria namely, correlation coefficient, root mean square error, Nash-Sutcliff coefficient and bias were used to evaluate and compare the performance of the models. The results showed that the performance of the models were satisfactory. Results showed that, the Bayesian neural network model is more efficient than the other models in estimation of minimum, mean and peak of runoff .

Keywords

Rainfall-Runoff

Bayesian neural network

Gene Expression programming

Support vector machine