Flat log-rate marginalization gives the multinomial likelihood

ID: flat-log-rate-marginalization-gives-the-multinomial-likelihood

For independent Poisson distribution cells with means , write and . Their likelihood, apart from count-factorial constants, is
An improper prior flat in has measure on . When , integration gives
so the remaining kernel in is , proportional to the multinomial likelihood. The gamma function integral is the exact reason for the equivalence.
This is proportionality of parameter kernels after an improper-prior calculation; it does not supply a normalized model evidence. For a zero group total, the baseline integral diverges at , so the same improper-prior marginalization is invalid even though the conditional multinomial likelihood is constant. A proper prior on the baseline generally changes the kernel. The Gaussian log-rate correction to the Poisson trick makes this additional factor explicit.

New to topics? Read the docs here!