Naser mohammad gholi mezerji - Faculty of Public Health, Department of Biostatistics, Hamedan University of Medical Sciences, Hamedan, Iran.
Majid sadeghifar - Faculty of Sciences, Department of Statistics, Bu-Ali Sina University, Hamedan, Iran.
Behnaz alafchi - Faculty of Public Health, Department of Biostatistics, Hamedan University of Medical Sciences, Hamedan, Iran.
Hemin shanazy - Faculty of Paramedical Sciences, Department of Biostatistics, Shahid Beheshti University of Medical Sciences,Tehran, Iran.

Objectives: Mortality rates and deaths from HIV/AIDS report annually and information for forecasting is not complete. Therefore, we have to choose a suitable method to fit the situation. According to the characteristics of this data, smallsize and imperfect, for analysis, we selected grey model first order one variable, GM(1,1). Materials and Methods: Data is collected and downloaded from Word Health Organization(WHO) reports in Iran at http://apps.who.int/gho/data/node.country.country- IRN. In data set, observed values between 2000 and 2013 used to fitting models and observed values in 2014 and 2015 used to evaluate the accuracy ofmodel’s predictions. To assess the efficiency of the model fitted and precise of predicted values, we used indexes of Forecast Absolute Error (FAE %), small error probability (P) and the proportion ofvariance (C). Results: Simulated results show that the accuracy of the model GM(1,1) to predict and forecast both data sets, but with a little suspicious for AIDS data, arelargely appropriate and reliable. Forecasting valuesat (2015,2020,2030) 2, 5 and 10 years later for neonatal mortality rate and number of deaths from HIV/AIDS are (9.25, 7.97, 6.22) and (62.58, 77.58, 110.98) respectively.Conclusions: In future studies and estimations, we suggest application of Grey Forecasting Models in other health information.

Neonatal mortality rate, HIV, forecasting Grey Model GM(1,1), forecasting