Derivative Calculation for Maximum Likelihood---I

• Let's do the same thing for maximum likelihood method.

  

• where are probability of our events.

• Assume are normalized as in Equation (25).

• This normalization is crucial but even when it does not seem to be true as in our subtle modifications in Section 4.3 with the ``generalized'' maximum likelihood method, one can easily modify the analysis, and show it is still valid.

• We get in normalized case,

  

• with