Research Article
BibTex RIS Cite

Comparison of Censored and Uncensored Poisson Regression Models

Year 2019, Volume: 9 Issue: 2, 359 - 376, 15.12.2019
https://doi.org/10.31466/kfbd.644229

Abstract

Poisson regression model is a regression model applied to events that occur in a certain period of time. In this model, the dependent variable consists of discrete count data. In this respect, it is a special type of regression models. Besides, Poisson regression model is one of the generalized linear models and is one of the most commonly used methods in applications. This model is applied for data showing equal spread. However, often the data sets do not meet the assumptions of the Poisson model. Sometimes the data set becomes censored for reasons such as illness, loss of the person or object being observed. If the dependent variable is censored, censored regression models are suitable for modeling over- or under-dispersed count data. In this study, Poisson regression models uncensored and censored are discussed. Both models were compared with IRR (incidence rate ratio), goodness of fit and information criteria. As a result of the study, it is shown that the censored Poisson regression model gives better results if the point to be censored is selected well.

References

  • Akaike, H. (1973). Information Theory and an Extension of the Maximum Likelihood Principle, InB.N. Petro and F. Csaki ed. 2nd International Symposium on Information Theory, 267-281.
  • Akın F. (2002). Kalitatif Tercih Modelleri Analizi, Bursa, Ekin Kitabevi.
  • Brännäs, K. (1992). Limited Dependent Poisson Regression. Statistician, 41: 413–423.
  • Caudill, S. B. ve Mixon Jr, F. G. (1995). Modeling Household Fertility Decisions: Estimation and Testing of Censored Regression Models for Count Data. Empirical Economics, 20(2): 183-196.
  • Hilbe, J. (2014). Modeling Cout Data, Cambridge University Press, 32 Avenue of the Americas, New York, NY 10013-2473, USA.
  • Hurvich, C.M. ve Tsai, C. (1989). Regression and Time Series Model Selection in Small Samples, Biometrika, 76: 297-307.
  • Famoye, F. (1993). Restricted Generalized Poisson Regression Model. Comm. Statist. Theory Methods, 22: 1335–1354.
  • Famoye, F., Wulu, J., ve Singh, K. P. (2004). On The Generalized Poisson Regression Model with an Application to Accident Data. Journal of Data Science, 2: 287-295.
  • Husain, M. ve Bagmar, S. H. (2015). Modeling Under-dispersed Count Data Using Generalized Poisson Regression Approach, Global Journal of Quantitative Science, 2(4): 22-29.
  • King, G. (1988). Statistical Models for Political Science Event Counts: Bias in Conventional Procedures and Evidence for the Exponential Poisson Regression Model, American Journal of Political Science, 32-(3): 838-863.
  • McQuarrie, A. D. ve Tsai, C.L. (1998). Regression and Time Series Model Selection, World Sciencetific.
  • Raciborski, R. (2011). Right-Censored Poisson Regression Model, The Stata Journal, 11(1): 95–105.
  • Saffari, S.E., Adnan, R. ve Greene, W. (2012). Parameter Estimation on Hurdle Poisson Regression Model with Censored Data. Jurnal Teknologi, 189-198.
  • StataCorp LLC. (2019). Sansürlü Poisson Regresyon Analizi, https://www.youtube.com/watch?v=6m_SXthPv1U
  • Sugiuna, N. (1978). Further Analysis of the Data by Akaike’s Information Criterion and the Finite Corrections, Communication in Statistics, Theory and Methods, 57: 13-26.
  • Terza, J. V. (1985). A Tobit-Type Estimator for the Censored Poisson Regression Model, Economics Letters 18: 361–365.
  • Wang, W. ve Famoye, F. (1997). Modeling Household Fertility Decisions with Generalized Poisson Regression. J. Population Econom, 10: 273–283.
  • Winkelmann, R. ve Zimmermann, K. F. (1995). Recent Developments in Count Data Modelling: Theory and Application. J. Econom, 9: 1–24.

Sansürlü Ve Sansürsüz Poisson Regresyon Modellerinin Karşılaştırılması

Year 2019, Volume: 9 Issue: 2, 359 - 376, 15.12.2019
https://doi.org/10.31466/kfbd.644229

Abstract

Poisson regresyon modeli, belli bir zaman periyodunda meydana gelen olaylara uygulanan bir regresyon modelidir. Bu modelde bağımlı değişken kesikli yani sayma verilerinden oluşur. Bu bakımdan regresyon modellerinin özel bir türüdür. Bunun yanı sıra Poisson regresyon modeli genelleştirilmiş doğrusal modeller arasında yer alır ve uygulamalarda en sık kullanılan yöntemlerden biridir. Bu model eşit yayılım gösteren veriler için uygulanmaktadır. Ancak çoğu zaman veri setleri Poisson modelinin varsayımlarını sağlamamaktadır. Bazen de veri seti hastalık, gözlemlenen kişinin ya da nesnenin kaybolması gibi nedenlerden dolayı sansürlü hale gelmektedir. Bu gibi bağımlı değişkenin sansürlü olması durumunda fazla veya az yayılım gösteren sayım verilerinin modellenmesi için sansürlü regresyon modelleri uygundur. Bu çalışmada sansürlü ve sansürsüz Poisson regresyon modelleri ele alınmıştır. Her iki model IRR (insidans oranı), uyum iyiliği ve bilgi kriterleri yardımıyla karşılaştırılmıştır. Çalışmanın sonucunda sansürleme yapılacak noktanın iyi seçilmesi durumunda sansürlü Poisson regresyon modelinin daha iyi sonuç verdiği gösterilmiştir.

References

  • Akaike, H. (1973). Information Theory and an Extension of the Maximum Likelihood Principle, InB.N. Petro and F. Csaki ed. 2nd International Symposium on Information Theory, 267-281.
  • Akın F. (2002). Kalitatif Tercih Modelleri Analizi, Bursa, Ekin Kitabevi.
  • Brännäs, K. (1992). Limited Dependent Poisson Regression. Statistician, 41: 413–423.
  • Caudill, S. B. ve Mixon Jr, F. G. (1995). Modeling Household Fertility Decisions: Estimation and Testing of Censored Regression Models for Count Data. Empirical Economics, 20(2): 183-196.
  • Hilbe, J. (2014). Modeling Cout Data, Cambridge University Press, 32 Avenue of the Americas, New York, NY 10013-2473, USA.
  • Hurvich, C.M. ve Tsai, C. (1989). Regression and Time Series Model Selection in Small Samples, Biometrika, 76: 297-307.
  • Famoye, F. (1993). Restricted Generalized Poisson Regression Model. Comm. Statist. Theory Methods, 22: 1335–1354.
  • Famoye, F., Wulu, J., ve Singh, K. P. (2004). On The Generalized Poisson Regression Model with an Application to Accident Data. Journal of Data Science, 2: 287-295.
  • Husain, M. ve Bagmar, S. H. (2015). Modeling Under-dispersed Count Data Using Generalized Poisson Regression Approach, Global Journal of Quantitative Science, 2(4): 22-29.
  • King, G. (1988). Statistical Models for Political Science Event Counts: Bias in Conventional Procedures and Evidence for the Exponential Poisson Regression Model, American Journal of Political Science, 32-(3): 838-863.
  • McQuarrie, A. D. ve Tsai, C.L. (1998). Regression and Time Series Model Selection, World Sciencetific.
  • Raciborski, R. (2011). Right-Censored Poisson Regression Model, The Stata Journal, 11(1): 95–105.
  • Saffari, S.E., Adnan, R. ve Greene, W. (2012). Parameter Estimation on Hurdle Poisson Regression Model with Censored Data. Jurnal Teknologi, 189-198.
  • StataCorp LLC. (2019). Sansürlü Poisson Regresyon Analizi, https://www.youtube.com/watch?v=6m_SXthPv1U
  • Sugiuna, N. (1978). Further Analysis of the Data by Akaike’s Information Criterion and the Finite Corrections, Communication in Statistics, Theory and Methods, 57: 13-26.
  • Terza, J. V. (1985). A Tobit-Type Estimator for the Censored Poisson Regression Model, Economics Letters 18: 361–365.
  • Wang, W. ve Famoye, F. (1997). Modeling Household Fertility Decisions with Generalized Poisson Regression. J. Population Econom, 10: 273–283.
  • Winkelmann, R. ve Zimmermann, K. F. (1995). Recent Developments in Count Data Modelling: Theory and Application. J. Econom, 9: 1–24.
There are 18 citations in total.

Details

Primary Language Turkish
Journal Section Articles
Authors

Öznur İşçi Güneri 0000-0003-3677-7121

Burcu Durmuş 0000-0002-0298-0802

Publication Date December 15, 2019
Published in Issue Year 2019 Volume: 9 Issue: 2

Cite

APA İşçi Güneri, Ö., & Durmuş, B. (2019). Sansürlü Ve Sansürsüz Poisson Regresyon Modellerinin Karşılaştırılması. Karadeniz Fen Bilimleri Dergisi, 9(2), 359-376. https://doi.org/10.31466/kfbd.644229