Quantile-based Fréchet half logistic distribution and its Bayesian inference
摘要
This paper introduces a novel quantile function formed by combining the quantile functions of the Fréchet and half logistic distributions, both widely used in modeling extreme values. The proposed distribution enhances flexibility in data modeling by integrating the distinct characteristics of these two distributions, thereby offering a broader framework to capture diverse data behaviors. We derive various quantile-based properties and reliability characteristics of this distribution, providing comprehensive insight into its structural properties. Estimation of model parameters is approached through quantile-based Bayesian inference, complemented by classical methods, including L-moments and percentile-based approaches. To demonstrate the effectiveness and versatility of the proposed model, we apply it to real-world datasets and conduct comparative analyses with established extreme value models.