Quasi-Bayes empirical Bayes: a sequential approach to the Poisson compound decision problem
A new statistical method called Quasi-Bayes empirical Bayes has been developed for the Poisson compound decision problem, particularly in streaming data scenarios. This sequential approach offers computational efficiency and constant per-observation costs as data accumulates. The method is supported by frequentist guarantees, including consistency and asymptotic optimality, and its performance has been validated through simulations and comparisons with existing benchmarks. AI