Modeling Skewed Dataset Using the Burr X-Exponential Distribution

Abstract

In this study, we propose an alternative distribution to exponential distribution called Burr-X Exponential distribution with two parameters. The exponential distribution, though widely used, has some lapses that include limited flexibility due to its single parametric nature, constant hazard rate, sensitivity to outliers, and poor fit to skewed datasets. Hence, the need to propose a robust distribution that can capture various data behaviours. The probability density function (PDF), cumulative distribution function (CDF), and reliability functions of the proposed (BX-E) model are derived. Other statistical properties, such as moments and quartile functions, are mathematically expressed. We estimate the parameter of the Burr-X Exponential (BX-E) model using the maximum likelihood estimation (MLE) method. We assess the efficiency of the (BX-E) model using simulated data and real-life applications, and the BX-E model outperforms existing distributions in terms of its minimum values of information criteria and goodness-of-fit tests.