We consider homogeneous scoring rules for selecting between Bayesian models for discrete data with possibly improper priors. Simulations indicate that, applied prequentially, the method will consistently select the true model.
科研之友
微信
新浪微博
Facebook
分享链接
