Next: Gaussian Limit of
Up: 3: Estimation of
Previous: Unnormalized Counter Experiment---III
- Now, we suppose t is no longer an unknown theoretical
parameter, but a known number.
- (b) is the form we found before.
- (a) can be regarded as
where 
is probability of
getting N (which should now be regarded as an experimental
observable) total counts in time t
- (
which is up to a constant,
the usual Poisson distribution).
- This requires some care to apply previous theorems as
probabilities must be normalized in proofs in Equation (14). However,
theorems do hold.
- Note Orear calls this normalized case as the extended maximum
likelihood method.
Geoffrey Fox, Northeast Parallel Architectures Center at Syracuse University, gcf@npac.syr.edu