The exp-\(G\) family of probability distributions (Q1930646): Difference between revisions

Let \(G(x;\theta)\) be the cumulative distribution function of a continuous or discrete random variable with \(\theta\) being a parameter. The authors introduce and study the two-parameter class of distributions \[ F^G_\lambda(x)= F_\lambda(G(x;\theta)), \] where, for \(x\in[0,1]\), \[ F_\lambda(x)=\begin{cases} (1- e^{-\lambda x})/(1- e^{-\lambda})& \text{if }\lambda\neq 0,\\ x&\text{if }\lambda= 0.\end{cases} \] They obtain several mathematical properties of \(F^G_\lambda(x)\), such as Kullback-Leibler divergence, Shannon entropy, moments and order statistics. They also estimate \(\lambda\) and \(\theta\) and present an application to real data.