An Empirical Bayes Optimal Discovery Procedure Based on Semiparametric Hierarchical Mixture Models

Figure 1

The nonparametric estimate for the differential component of the prior distribution of the prostate cancer data [38].