Diagnostic analysis for heterogeneous log-Birnbaum-Saunders regression models

In this paper, we introduce a class of heterogeneous log-Birnbaum–Saunders regression models where both the location parameter as well as the shape parameter can be modelled in terms of covariates. The maximum likelihood estimates of parameters are derived. In addition, diagnostic measures are studied based on the case-deletion approach and the local influence approach. Finally, one real example from the medical area is used to illustrate shape parameter modelling and the diagnostic methods.