DOI: 10.36347/sjavs.2014.v01i04.021

Pages: 260-264

In this study annual temperature data of Behshar during the period 1951-2008 is used to compare and predict the temperature using regression analysis and artificial neural network. After assessment of the statistical characteristics of annual temperature, assuming normality of the data was accepted by using the Kolmogorov-Smirnov method and drawing graphs. In Continue the regression methods for analyzing of time series component, R2 values for explaining performance of simple linear regression models and polynomial regression power and exponentially for predicting trend of time series compared with artificial neural network are used. The result of this study indicated that a direct and meaningful correlation exists in increasing trend of annual temperature in Behshar. Also the results show that artificial neural network is as precise as regression methods for simulation Behshar temperature in this study.