Karakaya Kadir, Rajitha C S, Sağlam Şule, Tashkandy Yusra A, Bakr M E, Muse Abdisalam Hassan, Kumar Anoop, Hussam Eslam, Gemeay Ahmed M
Department of Statistics, Faculty of Sciences, Selcuk University, Konya, Turkey.
Department of Mathematics, Amrita School of Physical Sciences, Amrita Vishwa Vidyapeetham, Coimbatore, 641112, India.
Sci Rep. 2024 Mar 27;14(1):7214. doi: 10.1038/s41598-024-57390-7.
This research commences a unit statistical model named power new power function distribution, exhibiting a thorough analysis of its complementary properties. We investigate the advantages of the new model, and some fundamental distributional properties are derived. The study aims to improve insight and application by presenting quantitative and qualitative perceptions. To estimate the three unknown parameters of the model, we carefully examine various methods: the maximum likelihood, least squares, weighted least squares, Anderson-Darling, and Cramér-von Mises. Through a Monte Carlo simulation experiment, we quantitatively evaluate the effectiveness of these estimation methods, extending a robust evaluation framework. A unique part of this research lies in developing a novel regressive analysis based on the proposed distribution. The application of this analysis reveals new viewpoints and improves the benefit of the model in practical situations. As the emphasis of the study is primarily on practical applications, the viability of the proposed model is assessed through the analysis of real datasets sourced from diverse fields.
本研究开始了一个名为幂新幂函数分布的单位统计模型，对其互补性质进行了全面分析。我们研究了新模型的优点，并推导了一些基本的分布性质。该研究旨在通过呈现定量和定性的认识来提高洞察力和应用。为了估计模型的三个未知参数，我们仔细研究了各种方法：最大似然法、最小二乘法、加权最小二乘法、安德森-达林法和克拉默-冯米塞斯法。通过蒙特卡罗模拟实验，我们定量评估了这些估计方法的有效性，扩展了一个稳健的评估框架。本研究的一个独特之处在于基于所提出的分布开发了一种新颖的回归分析。这种分析的应用揭示了新的观点，并提高了模型在实际情况下的效益。由于研究的重点主要是实际应用，因此通过对来自不同领域的真实数据集的分析来评估所提出模型的可行性。
