Abstract: Suppose that an observed count is of the form , where the background and the signal are independent Poisson random variables with parameters and , is known, but is not. The model arises in astronomy and high-energy physics, and some recent articles have suggested conditioning on the observed bound for ; that is, if is observed, then the suggestion is to base the inference on the conditional distribution of given . This suggestion is used here to derive an estimator of the signal, and the estimator is shown to be admissible and minimax.
Key words and phrases: Admissible, Confidence and credible interval, coverage probability, mean square error, minimax, signal plus background.