Performance Evaluation of Neural Network Models to Drought Events Prediction in Yazd Station

Nowadays, computational intelligence models have shown high ability on prediction of time series. The main purpose of this paper is to evaluate the performance of six artificial neural network based models for simulation of precipitation and subsequent drought condition in Yazd station and identifying the best one. In order to evaluate the performance of models two criteria: Root Mean Square Error (RMSE) and Correlation Coefficient (R) were used. The results showed that the dynamic structures of artificial neural network including recurrent network (RN) and Time Lag Recurrent Network (TLRN) showed better performance with R=0.77 and 0.78, respectively. Comparing the performance of RN and TLRN models indicated that, however both of the models had the similar performance, but the TLRN network was the best one.