Using Bayesian Weighted Method to Estimate the Parameters of Qualitative Regression Depending on Poisson distribution "A comparative Study"


       In this paper, we are estimating Weights for parameters models of qualitative regression using Weighted Least Square Method (WLS),which has estimate through inverted variance of the distribution, by two Methods, first one (Classical Method) and the second (Sequential Bayesian Method), when the depended variable is binary variable, and the observations which represented the samples that we take for study distributed non-normality distribution as Poisson distribution, and related to communications have the problem of non-homogeneity variance of random error.

For treatment the problem of heterogeneity of the random error variance we estimate weights used in the way (wls) by two Methods and estimate qualitative regression parameters then comparison between results of Classical Method with Sequential Bayesian Method through some statistic measurements as (MSE, R2, F-test).

Dealt with practical side of the relationship between miscarriages of women as dependent variables and two independent variables, Age of women at marriage and Children ever born ,The data has been taken from Multiple Indicators Cluster (MICS4) survey 2011, the No. of observations was (8024) women where these observations, we noted by using Sequential Bayesian Method we get minimum (MSE) less than the Classical Method in the other side the results of (R-square and F-test) in the Classical Method is better than the Sequential Bayesian Method, We analyzed the data depending on programs written in MATLAB Language and statistics program (SPSS).


Annette, J.Dobson. (2002) An Introduction to Generalized Linear Models. 2ndEd.,Chapman&Hall/CRC,U.S.A,Florida. pp. (15,52,143).
Box & Tiao, G.C. (1973) Bayesian Inference In Statistical Analysis. Addison-Wesley Publishing Company, California, London, pp.(10 -12).
Chtterjee, Samprit & Bertam, Price.(1997) Regression Analysis by Example. John Wily and Sons, New York.
Draper, N.R. & Smith, H. (1981) Applied Regression Analysis. Second Edition, John Wiley &Sons, New York, pp.(108-116).
Draper, N. R.& Smith, H. (1998) Applied Regression Analysis. Wiley-Inter science.
Freund, Rudolf,J. And William, Wilson.(1997) Statistical Methods. Academic Press, New York.
Harrison P. J. & Stevens C. F. (1971) A Bayesian Approach to Short – Term Forecasting. Operation Research, Vol. 22, No.2, pp. (341-362).
Kleinbaum, David G. & Nizam, Azhar.(1998) Applied Regression Analysis and other Multivariable Methods.3rd Ed, An International Thomson Publishing, USA.
Kypraios, Theo (2013) Introduction to Bayesian Statistics (And some Computational Methods).MSc in Applied Bioinformatics Craneld University.p.27
Lomax, Richard G.(2007) Statistical Concepts: Second Course. p.10
Maddala,G.S. (2003) Introduction to Econometrics. 3rd ed, John Wiley and Sons, New York.
Massey William. A, Parker Geraldine.A,Whitt Ward.(1993) Poisson Estimating the Parameters of a Nonhomogeneous Poisson Process with Linear Rate. AT &T Bell Laboratories Murray Hill, pp.(6,10,42 ).
Rawlings, John,O.,Pantula, Sastry, G., Dickey, David A.(1998) Applied Regression Analysis: A Research Tool. Second Edition, Springer, pp.(76-79).
Uusipaikka, Esa. (2008) Confidence Intervals in Generalized Regression Models. University of Turku Turun Yliopisto , Finland, p.217.
Weisberg Sanford. (2005) Applied Linear Regression. 3rd Ed, John Wiley & Sons, New Jersey ,p.97.
How to Cite
“Using Bayesian Weighted Method to Estimate the Parameters of Qualitative Regression Depending on Poisson distribution "A comparative Study"” (2016) Zanco Journal of Pure and Applied Sciences, 28(5), pp. 41-52. doi: 10.21271/zjpas.v28i5.556.