Research Article
BibTex RIS Cite

Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family

Year 2025, Volume: 37 Issue: 4, 470 - 479, 23.12.2025
https://doi.org/10.7240/jeps.1815577

Abstract

This study proposes a new family of continuous probability distributions, called the Continuous Bernoulli-G (CB-G), which is constructed using the T–X framework by adopting the Continuous Bernoulli (CB) distribution as the generator. Defined on a positive and continuous interval, this family provides a flexible modeling framework by combining the structural properties of the CB distribution with various baseline distributions. The primary motivation behind developing this new family is to generate alternative distributions that are particularly suitable for modeling pixel intensities in grayscale image analysis.Fundamental statistical properties of the proposed family—such as the probability density and cumulative distribution functions, quantile function, moments, entropy, reliability measures, and maximum likelihood estimation—are derived. Three special submodels of the CB-G family, based respectively on the Beta, Uniform, and Weibull distributions, are examined in detail due to their mathematical tractability and practical relevance. The proposed models are applied to grayscale image data extracted from the CIFAR-100 dataset, and their performance is evaluated using log-likelihood values and information criteria (such as AIC). The results show that, in particular, the CB-B and CB-W distributions outperform the classical Beta and CB distributions in modeling pixel intensity distributionsThis study demonstrates the potential of the proposed family to contribute to real-world data modeling problems and provides a foundation for a wide range of future theoretical and applied research efforts.

References

  • Theis, L., van den Oord, A., & Bethge, M. (2016). A note on the evaluation of generative models. arXiv preprint arXiv:1511.01844.
  • Kingma, D. P., & Welling, M. (2014). Auto-encoding variational Bayes. International Conference on Learning Representations (ICLR).
  • Loaiza-Ganem, G., & Cunningham, J. P. (2019). The continuous Bernoulli: Fixing a pervasive error in variational autoencoders. Advances in Neural Information Processing Systems (NeurIPS).
  • Wang, K.-S., & Lee, M.-Y. (2020). Continuous Bernoulli distribution—simulator and test statistic. Ji-Tong Co., Ltd.
  • Alzaatreh, A., Lee, C., & Famoye, F. (2013). A new method for generating families of continuous distributions. Metron, 71(1), 63–79.
  • Mudholkar, G. S., & Hutson, A. D. (1996). The exponentiated Weibull family: Some properties and a flood data application. Communications in Statistics – Theory and Methods, 25(12), 3059–3083.
  • Marshall, A. W., & Olkin, I. (1997). A new method for adding a parameter to a family of distributions with application to the exponential and Weibull families. Biometrika, 84(3), 641–652.
  • Eugene, N., Lee, C., & Famoye, F. (2002). Beta-normal distribution and its applications. Communications in Statistics – Theory and Methods, 31(4), 497–512.
  • Zografos, K., & Balakrishnan, N. (2009). On families of beta- and generalized gamma-generated distributions and associated inference. Statistical Methodology, 6(4), 344–362.
  • Bourguignon, M., Silva, R. B., & Cordeiro, G. M. (2014). The Weibull–G family of probability distributions. Journal of Data Science, 12(1), 53–68.
  • Gomes-Silva, F., Percontini, A., de Brito, E., Ramos, M. W., Venâncio, R., & Cordeiro, G. M. (2017). The Odd Lindley-G family of distributions. Austrian Journal of Statistics, 46(1), 65–87.
  • Cakmakyapan, S., & Ozel, G. (2017). The Lindley family of distributions: Properties and applications. Hacettepe Journal of Mathematics and Statistics, 46(6), 1113–1137.
  • Elbatal, I., Ozel, G., & Cakmakyapan, S. (2022). Odd extended exponential-G family: Properties and application on earthquake data. Journal of Statistics and Management Systems, 25(8), 1751–1765.
  • Kenney, J. F., & Keeping, E. S. (1962). Mathematics of statistics (Part 1, 3rd ed.). Princeton, NJ: D. Van Nostrand.
  • Moors, J. J. A. (1988). A quantile alternative for kurtosis. Journal of the Royal Statistical Society: Series D (The Statistician), 37(1), 25–32.
  • Soriani, F. (2021). CIFAR-100 - Image classification dataset [Data set]. Kaggle. https://www.kaggle.com/datasets/fedesoriano/cifar100

Yeni Bir Genelleştirilmiş Olasılık Dağılım Ailesi Kullanılarak Piksel Yoğunluklarının İstatistiksel Modellemesi

Year 2025, Volume: 37 Issue: 4, 470 - 479, 23.12.2025
https://doi.org/10.7240/jeps.1815577

Abstract

Bu çalışmada, Sürekli Bernoulli-G (CB-G) adı verilen yeni bir sürekli olasılık dağılımları ailesi önerilmektedir. Bu aile, Sürekli Bernoulli (CB) dağılımını üreteç (generator) olarak kullanarak T–X çerçevesi aracılığıyla oluşturulmuştur. Pozitif ve sürekli bir aralıkta tanımlanan bu aile, CB dağılımının yapısal özelliklerini çeşitli temel (baseline) dağılımlarla birleştirerek esnek bir modelleme çerçevesi sunmaktadır. Bu yeni ailenin geliştirilmesindeki temel motivasyon, özellikle gri tonlu görüntü analizlerinde piksel yoğunluklarını modellemeye uygun alternatif dağılımlar elde etmektir. Önerilen ailenin olasılık yoğunluk ve kümülatif dağılım fonksiyonları, kantil fonksiyonu, momentleri, entropisi, güvenilirlik ölçüleri ve maksimum olabilirlik tahminleri (MLE) gibi temel istatistiksel özellikleri türetilmiştir. CB-G ailesinin üç özel alt modeli, sırasıyla Beta, Uniform (Düzgün) ve Weibull dağılımlarına dayalı olanlar, matematiksel olarak uygulanabilirlikleri ve pratik önemleri nedeniyle ayrıntılı biçimde incelenmiştir. Önerilen modeller, CIFAR-100 veri kümesinden elde edilen gri tonlu görüntü verilerine uygulanmış ve performansları log-olasılık değerleri ile AIC gibi bilgi ölçütleri kullanılarak değerlendirilmiştir. Sonuçlar, özellikle CB-B ve CB-W dağılımlarının, piksel yoğunluğu dağılımlarını modellemede klasik Beta ve CB dağılımlarından daha başarılı olduğunu göstermektedir. Bu çalışma, önerilen genelleştirilmiş dağılım ailesinin gerçek dünya veri modelleme problemlerine olası katkılarını ortaya koymakta ve geniş bir yelpazede yürütülebilecek teorik ve uygulamalı araştırmalar için bir temel oluşturmaktadır.

References

  • Theis, L., van den Oord, A., & Bethge, M. (2016). A note on the evaluation of generative models. arXiv preprint arXiv:1511.01844.
  • Kingma, D. P., & Welling, M. (2014). Auto-encoding variational Bayes. International Conference on Learning Representations (ICLR).
  • Loaiza-Ganem, G., & Cunningham, J. P. (2019). The continuous Bernoulli: Fixing a pervasive error in variational autoencoders. Advances in Neural Information Processing Systems (NeurIPS).
  • Wang, K.-S., & Lee, M.-Y. (2020). Continuous Bernoulli distribution—simulator and test statistic. Ji-Tong Co., Ltd.
  • Alzaatreh, A., Lee, C., & Famoye, F. (2013). A new method for generating families of continuous distributions. Metron, 71(1), 63–79.
  • Mudholkar, G. S., & Hutson, A. D. (1996). The exponentiated Weibull family: Some properties and a flood data application. Communications in Statistics – Theory and Methods, 25(12), 3059–3083.
  • Marshall, A. W., & Olkin, I. (1997). A new method for adding a parameter to a family of distributions with application to the exponential and Weibull families. Biometrika, 84(3), 641–652.
  • Eugene, N., Lee, C., & Famoye, F. (2002). Beta-normal distribution and its applications. Communications in Statistics – Theory and Methods, 31(4), 497–512.
  • Zografos, K., & Balakrishnan, N. (2009). On families of beta- and generalized gamma-generated distributions and associated inference. Statistical Methodology, 6(4), 344–362.
  • Bourguignon, M., Silva, R. B., & Cordeiro, G. M. (2014). The Weibull–G family of probability distributions. Journal of Data Science, 12(1), 53–68.
  • Gomes-Silva, F., Percontini, A., de Brito, E., Ramos, M. W., Venâncio, R., & Cordeiro, G. M. (2017). The Odd Lindley-G family of distributions. Austrian Journal of Statistics, 46(1), 65–87.
  • Cakmakyapan, S., & Ozel, G. (2017). The Lindley family of distributions: Properties and applications. Hacettepe Journal of Mathematics and Statistics, 46(6), 1113–1137.
  • Elbatal, I., Ozel, G., & Cakmakyapan, S. (2022). Odd extended exponential-G family: Properties and application on earthquake data. Journal of Statistics and Management Systems, 25(8), 1751–1765.
  • Kenney, J. F., & Keeping, E. S. (1962). Mathematics of statistics (Part 1, 3rd ed.). Princeton, NJ: D. Van Nostrand.
  • Moors, J. J. A. (1988). A quantile alternative for kurtosis. Journal of the Royal Statistical Society: Series D (The Statistician), 37(1), 25–32.
  • Soriani, F. (2021). CIFAR-100 - Image classification dataset [Data set]. Kaggle. https://www.kaggle.com/datasets/fedesoriano/cifar100
There are 16 citations in total.

Details

Primary Language English
Subjects Applied Statistics
Journal Section Research Article
Authors

Selen Çakmakyapan 0000-0002-1878-2181

Submission Date November 2, 2025
Acceptance Date December 16, 2025
Publication Date December 23, 2025
Published in Issue Year 2025 Volume: 37 Issue: 4

Cite

APA Çakmakyapan, S. (2025). Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family. International Journal of Advances in Engineering and Pure Sciences, 37(4), 470-479. https://doi.org/10.7240/jeps.1815577
AMA Çakmakyapan S. Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family. JEPS. December 2025;37(4):470-479. doi:10.7240/jeps.1815577
Chicago Çakmakyapan, Selen. “Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family”. International Journal of Advances in Engineering and Pure Sciences 37, no. 4 (December 2025): 470-79. https://doi.org/10.7240/jeps.1815577.
EndNote Çakmakyapan S (December 1, 2025) Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family. International Journal of Advances in Engineering and Pure Sciences 37 4 470–479.
IEEE S. Çakmakyapan, “Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family”, JEPS, vol. 37, no. 4, pp. 470–479, 2025, doi: 10.7240/jeps.1815577.
ISNAD Çakmakyapan, Selen. “Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family”. International Journal of Advances in Engineering and Pure Sciences 37/4 (December2025), 470-479. https://doi.org/10.7240/jeps.1815577.
JAMA Çakmakyapan S. Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family. JEPS. 2025;37:470–479.
MLA Çakmakyapan, Selen. “Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family”. International Journal of Advances in Engineering and Pure Sciences, vol. 37, no. 4, 2025, pp. 470-9, doi:10.7240/jeps.1815577.
Vancouver Çakmakyapan S. Statistical Modeling of Pixel Intensities Using a Novel Generalized Probability Distribution Family. JEPS. 2025;37(4):470-9.