Empirical Bayes (Robbins, 1956; Efron and Morris, 1977; Casella, 1985) refers to methods which exploit such forms of indirect information (Efron, 2010) to reliably deliver highly individualized inferences—here, the test results for the other n patients provide indirect information about your disease status.

(Liu & Meng, 2015) Is Empirical Bayes related to §meta_learning then?

Bibliography

Liu, K., & Meng, X. (2015), There is individualized treatment. why not individualized inference?, CoRR. ↩