Abstract
Panel count data have been extensively discussed in the literature. In general, the existing approaches in modeling panel count data usually assume an exponential form for the dependence of the conditional mean function on covariate variables. However, this assumption may be violated in practice. A more flexible panel count data model with an unknown link function is proposed, and a local logarithm partial likelihood function is formed for the estimation. A two-step iterative algorithm is employed to estimate the unknown link function and covariate effects. Furthermore, the baseline function is obtained by Breslow estimation. Asymptotic properties are derived under some mild conditions. Some numerical simulations and an application of bladder cancer are carried out to confirm and assess the performance of the proposed model and approach.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call