In: Statistics and Probability
Question:
(Bayesian) Suppose that X is Poisson(λ + 1), and the prior distribution of λ is binomial(2,1/3).
(a) Find the Bayesian estimate of λ for mean square loss based on the single observation X, if X = 1.
(b) Find the Bayesian estimate of λ for mean square loss based on the single observation X, if X = 2
Hints:
Because of its prior distribution, λ can take only three values, 0,1,2.
Don’t expect its posterior distribution to be any distribution we have seen before;
You will have to compute (numerically) the posterior probabilities of the three possible values.