Maximum-minimum temperature prediction using fuzzy random auto-regression time series model

Many models have been suggested to predict the weather and temperature data. Most of them used the single point data in building prediction equations. Besides that, the randomness, the vagueness and possibility of the temperature data are also not much concerned. In this paper, we introduce the minimum-maximum procedure for daily temperature modeling based on fuzzy random auto-regression time series. The proposed procedure was able to cover the variability of the temperature in nature. The result showed that mean square error of proposed model is smaller than the existing models.