Statistical Analysis & Asthmatic Patients in Sulaimaniyah Governorate in the tuber-closes center Dr.Mohammad M.Faqe Hussain(1) University of Sulaimani School of Administration and Economics Statistical Department faqi_zanko@yahoo.co.uk Asst. Lecturer Roqia Abdulkadir Jamel(2) Presidency of Slemani Polytechnic University Technical Institute of Slemani Statistics and informatics department Roqia_Abdulkadir@yahoo.com Abstract This paper is study the people who was affected with asthma diseases during the period (2005-2011). Asthma is a disease affecting the airways that carry air to and from your lungs. People who suffer from this chronic condition (long-lasting or recurrent) are said to be asthmatic. The inside walls of an asthmatic's airways are swollen or inflamed. This swelling or inflammation makes the airways extremely sensitive to irritations and increases your susceptibility to an allergic reaction. in this study applied two models, the first model is single exponential smoothing model(SESM) and second model is the double exponential smoothing model (SDEM) ,and after that it compare between the two models depending on the measures (MAPE, MAD, MSD), and after the process of comparison the double exponential smoothing model was selected ". Depending on the double exponential smoothing model the confidence limit and prediction of numbers of people who is affected with these diseases are obtained subsequent the years 2012 and 2013. In the final of this study the number of male and female affected with asthma disease are increasing in the years 2012 and 2013 for both genders .The study recommend that the Ministry of Health must be more attention and care, for the patients who affected with this type of the disease in the sulaymaniyah governorate and then supply of medicines and open more health centers for the purpose of their care and treatments. Keywords: Method of Analysis, the model, Forecasting Performance Measures, Analysis the data 1 Method of Analysis [2][4] In this paper, two different types of univariate forecasting model were developed on the same data set for the purpose of finding the best fit models. These models are Single Exponential Model, Double Exponential Model. The application of univariate models (Single Exponential Model, Double Exponential Model) as forecasting provide the advantage of simplicity, less costly to develop and easier to understand. Each model type has its unique characteristics which define its suitability to be fitted to that data. Data collected were analyzed using Microsoft Office Excel and MINITAB. 2 The Model 2-1 Single Exponential Smoothing Model [1][3][6] Single exponential smoothing model is the simplest model in the group of the exponential smoothing techniques. The model requires only one parameter, which is constant, α, to generate the fitted values and forecast. In this model, the forecast for the next and all subsequent periods are determined by adjusting the current period forecast by a portion of the difference between the current forecast and current actual value. Hence, mathematically the model can represented as: πΉπ‘+π = πΌπ¦π‘ + (1 − πΌ)πΉπ‘ ………..(1) Where: πΉπ‘+π = single exponential smoothed value in period π‘ + π π¦π‘ = the actual value at time period π‘ πΌ = unknown smoothing constant to be determined. The πΌ value is between 0 and 1. πΉπ‘ = the forecast made in period π‘ The best parameter value, α can be determined by using solver facility that available in Excel software based on the estimation part of the data. The best value of α is defined based on the smallest value of MSD 2-1 Double Exponential Smoothing Model [1][3] [6] This model is useful for series that exhibits a linear trend characteristic, the main advantage of double exponential smoothing method over single exponential is it ability to generate multiple-ahead-forecast. Let: St be the exponentially smoothed value of Yt at time t S’t be the double exponentially smoothed value of Yt at time t Computes the single exponentially smoothed value: St = αYt + (1 – α)St-1 ……………………….(2) Computes the double exponentially smoothed value: S’t = αSt + (1 – α)S’t-1 ……………………….(3) Next, ππ‘ = 2St – S’t And, computes the adjustment factors: ππ‘ = α 1−α (ππ‘ − π ′ π‘ ) Forecast for m-step-ahead are computed using the equation: πΉπ‘+π = ππ‘ + ππ‘ π ………………….(4) As in exponential modeling, the main difficulty encountered when using this method is the determination of the size α. As usual, the criterion is to choose α such that the MSD is minimum. Again, solver used to solve this problem [8] . 3 Forecasting Performance Measures [5] [7] ο· Mean Forecast Error (MFE or Bias): Measures average deviation of forecast from actual. MFE ο½ ο· 1 n ο₯ ( At ο Ft ) n t ο½1 ……………(5) Mean Absolute Deviation (MAD): Measures average absolute deviation of forecast from actual MAD ο½ ο· 1 n ο₯ At ο Ft n t ο½1 ………………(6) Mean Absolute Percentage Error (MAPE): Measures absolute error as a percentage of the forecast. MAPE ο½ ο· 100 n At ο Ft ο₯ n t ο½1 At ……………..(7) Standard Squared Error (MSE): Measures variance of forecast error MSE ο½ 1 n ( At ο Ft ) 2 ο₯ n t ο½1 ………………(8) 4 Analysis the data: In this research, asthma disease data was used. The data was obtained from the hospital in Sulaimaniyah governorate. A monthly data set for 7 years period (2005 to 2011) is used and the data are shown in the figure below: 2500 2000 1500 Male 1000 Female 500 0 2005 2006 2007 2008 2009 2010 2011 Figure (1): Show the numbers of people infected with the asthma disease during the year (2005-2011) First: We draw the curve of the number of people infected with asthma during the year (20052011) .The horizontal axis for time and the vertical axis value of numbers of (male, female) patients with asthma disease. 140 120 Male 100 80 60 40 20 0 Index 10 20 30 40 50 60 70 80 Figure (2) : The time series of the number of male patients infected with asthma disease during the years (2005-2011) Female 250 150 50 Index 10 20 30 40 50 60 70 80 Figure (3): The time series of the number of female patients infected with asthma disease during the years (2005-2011) Second: We are finding the values of the measurements (MAPE,MAD,MSD ) for the two models below and according to the data the males infected with this type of disease: Table (1): show the measures of the two models Model α γ MAPE Single Exponential 0.1 - 44.211 16.136 447.846 MAD MSD Double Exponential 0.2 0.2 42.001 15.137 355.747 140 120 Male 100 80 60 40 20 0 1 8 16 24 32 40 48 Index 56 64 72 80 Figure (4): Show Smoothing Plot for Male (Single Exponential Smoothing) 140 120 Male 100 80 60 40 20 0 1 8 16 24 32 40 48 Index 56 64 72 80 Figure (5): Show Smoothing Plot for Male (Double Exponential Smoothing) Through the table (1) we comparison between the two models by the measures of (MAPE, MAD, MSD).From the measures we note that the best model is double exponential smoothing model because the values of measures (MAPE, MAD, MSD) be less than the measures of single exponential smoothing model. Table (2) : Show the prediction and confidence interval of the numbers of male patients infected with asthma disease during the year (2012-2013) Year Period 1 2 3 4 5 6 2012 7 8 9 10 11 12 Forecasting 57.65 Lower Upper 20.57 61.49 23.63 65.33 26.64 69.16 29.59 73.00 32.48 76.84 35.33 80.68 38.14 84.51 40.90 88.35 43.62 92.19 46.31 96.02 48.96 99.86 51.58 Year Period Forecasting 1 94.74 2 99.35 3 104.02 4 108.74 5 113.52 118.34 6 2013 7 123.21 8 128.13 9 133.08 10 138.07 11 143.09 12 148.14 Lower Upper 103.70 54.18 153.22 107.53 56.74 158.33 111.37 59.29 163.46 115.21 61.81 168.61 119.05 64.30 173.79 122.88 66.78 178.98 126.72 69.24 184.20 130.56 71.69 189.43 134.39 74.12 194.67 138.23 76.53 199.94 142.07 78.93 205.21 145.91 81.31 210.50 Depending on the table above we note that the number of the people who infected with the asthma disease in the governorate of Sulaymaniyah of the male patients are increased over the years 2012, 2013 and this reason is due to lack of health awareness among the people and the people who infected with this type of the disease are not revision a specialist doctor for the purpose of taking the necessary treatment. 200 Male 150 100 50 0 1 11 22 33 44 55 Index 66 77 88 99 Figure (6): Show the prediction of the number of male patients infected with the asthma diseases Third: We are finding the values of the measurements (MAPE,MAD,MSD ) for the two models below and according to the data the female infected with this type of disease: Table (3): Show the measures of the two models Model α γ MAPE MAD Single Exponential 0.1 - 30.01 29.11 1477.37 28.62 26.51 1112.35 Double Exponential 0.2 0.2 MSD 250 Female 200 150 100 50 1 8 16 24 32 40 48 Index 56 64 72 80 Figure (7): Show Smoothing Plot for Female (Single Exponential Smoothing) 250 Female 200 150 100 50 1 8 16 24 32 40 48 Index 56 64 72 80 Figure (8): Show Smoothing Plot for Female (Double Exponential Smoothing) Through the table (3) we comparison between the two models by the measures of (MAPE, MAD, MSD).From the measures we note that the best model is double exponential smoothing model because the values of measures (MAPE, MAD, MSD) be less than the measures of single exponential smoothing model. Table (4): Show the prediction and confidence interval of the numbers of Female patients infected with asthma disease during the year (2012-2013) Year Period Forecasting 1 123.40 58.45 2 130.59 64.28 3 Lower 137.77 4 70.01 144.96 75.65 Upper Year Period Forecasting Lower Upper 188.35 1 209.66 122.93 296.39 196.89 2 216.85 127.89 305.80 205.54 3 224.03 132.81 315.26 214.28 4 231.22 137.69 324.75 5 152.15 81.19 223.11 5 238.41 142.54 334.29 6 159.34 86.65 232.03 6 245.60 147.34 343.85 241.03 7 252.79 152.12 353.45 2012 7 166.53 92.02 2013 8 173.72 97.33 250.10 8 259.98 156.87 363.08 9 180.90 102.57 259.24 9 267.16 161.59 372.74 268.45 10 274.35 166.28 382.42 281.54 170.95 392.13 288.73 175.60 401.86 10 188.09 107.74 11 195.28 112.85 277.71 11 12 202.47 117.91 287.02 12 Depending on the table above we note that the number of the people who infected with the asthma disease in the governorate of Sulaymaniyah of the female patients are increased over the years 2012, 2013 and this reason is due to lack of health awareness among the people and the people who infected with this type of the disease are not revision a specialist doctor for the purpose of taking the necessary treatment. 400 Female 300 200 100 0 1 11 22 33 44 55 Index 66 77 88 99 Figure (9): Show the prediction of the number of female patients infected with the asthma diseases 5 Conclusions & Recommendations: 5-1 Conclusions Through this study, we get a set of conclusions and must be necessary mentioned as follows: 1) Through the data analysis we note that the appropriate model to predict the number of males patients infected with asthma disease is double exponential smoothing model 2) The appropriate model to predict the number of female patients infected with asthma is double exponential smoothing model 3) From the data analysis we see that the numbers of male and female patient infected with asthma disease are increases during the year (2012-2013). 5-2 Recommendations: There is a set of recommendations for this study and these recommendations are: - : 1) People with this type of disease must take care of them by the government and open more health centers for a dangerous disease that affects the respiratory system to humans. 2) Provides more data and information by the health center for researchers to conduct more research in this type of disease. 3) Using statistical measurements more accurate to analyze this type of disease. References [1] A. Pankratz,(1983) . Forecasting with univariate Box-Jenkins models, John Wiley, John Wiley & Sons New York. [2] Bowerman, B.L., O’Connell, R.T. & Koehler, A.B. (2005). Forecasting, Time Series, and Regression An Applied Approach. Ed: 4. United States of America: Thomson Brooks/Cole. [3] C. Chatfield, B. Koehler, K. Ord, and D. Snyder (2001). A new look at models for exponential smoothing, The Statistician 50, no. 2, 147-159. [4] Hanke, J.E. & Wichern, D.W. (2005). Business Forecasting. Ed: 8. United States of America: Pearson Prentice Hall. [5] Krajewski,L.J.,Ritzman,L.R.,,(1993).Operations Management, Strategy and Analysis, Addision -Wesley Publishing Company,Inc.,P.436 [6] Mohd Alias Lazim (2007). Introductory Business Forecasting. A Practical Approach. Shah Alam: Pusat Penerbitan Universiti, Universiti Teknologi Mara. [7] Makridakis,S.,Wheelwright,S.C.,(1989).Forecasting Methods for Management, John Wiley & Sons New York,P.3. [8] Wilson, J.H & Keating, B. (2002). Business Forecasting, Ed:4. New York: McGrawHill.