Mixture of Pathway Weibull Model and Mellin Convolutions

The eciency of the pathway idea in modeling heterogeneous real-life data is illustrated in this paper. Usually mixture models are used for modeling heterogeneous data. Here we introduce a nite mixture of two pathway Weibull models, denoted by WM_q (or _qWM or W_qM). Properties of this new model are examined and identiability is proved. Some important special cases of WM_q are given. Stress-strength reliability is found by using the Mellin convolution technique. With the help of a real heterogeneous data set, it is shown that the proposed model ts the data better than all other popular models in the literature.