Regression Modeling Using Generalized Pareto‐IV Distribution With Application to Household Budget Survey

Abstract

The study introduced the new regression model by the T‐X family technique of Type I half‐logistic and Pareto IV distribution (TIHLPIV) that had to solve the weaknesses of mixed methods, which are limited to applicability, data requirements, and computational complexity. Data transformation was needed for generalization (G‐family) of combined distributions, and the T‐X technique allowed flexibility, risk assessment, environmental studies, quality control, and scientific research in modeling different kinds of data and covering a wider range of real‐life applications such as survival and lifetime data, hydrology, economics (income inequality), and engineering. The general objective of the study was to introduce new regression modeling using generalized Pareto IV distribution (TIHLPIV) with application to the Household Budget Survey (HBS) for the years 2017‐18 on the Tanzania Mainland. According to empirical results, it is revealed that the proposed model gives better results than other existing models.