Abstract
Logistic regression for binary response variable is often used many fields of study for instance clinical applications humanitarian and social issues. In this paper, Bayesian estimation approach is introduce to estimate the unknown link function and the coefficient vector in the semiparametric logistic regression. The normal distribution prior was considered to the coefficient vector and Gaussian process prior was set for the unknown link function. Bayesian hierarchical model was developed for the single index logistic regression model. MCMC algorithm was adopted for posterior inference. To compare our proposed method BSLR with the existing methods real data and two simulation examples are considered. We have conclude that our proposed method do well.
