Okay so I have been on this for like 4 hours. I am at my wit's end. (Side note: I honestly feel the notes should be a bit more explicit instead of just assuming we are all Harvard/MIT level genius.)
I think I understand a bit, but there is a bit more confusion.
Let me take it from the top:
X represents the total number of claims/ year from a Low risk policy (LRP). X~Poi(mu).
Similarly, Y represents the total number of claims/year from a HRP. Y~Poi(2 *mu).
Now there are 48 claims from 15 LRPs in a year. So lets say X = 48. So X ~ Poi(15*mu).
Similarly, there are 59 claims from 10 HRPs in a year. Therefore, Y~ Poi(10*2*mu).
Now,
L(mu)= (Prod 1st Policy to 15th Policy) [P(X=xi)] * (Prod 1st Policy to 10th Policy) [P(Y=yi)]
x(i) represents the number of claims for the i(th) policy. So sum of x(1) to x(15) = 48 (as there are 48 claims from 15 policies). Since now we are back to taking about an individual claim, the P(X=x(i))=Poisson prob func with mu as mean.
From there it is easy to follow the calculations.
I really hope before they jump into the solution, they spent some ink defining the variables.
If my thought process is wrong, please correct me.
Last edited by a moderator: Jan 11, 2016