First, concerning truncation, Q.B20 page 367
For type Y policies they are truncated from above and there are 50
losses less than k and a total of 75 losses that exceed k and have
been recorded. they ask you for the ml estimate of theta.
the answer they give you is as follows:
L(theta)=Product from 1 to 50 of f(yi;theta) times [1-F(k;theta)]^75
Why isn't Product from1 to 50 of[ f(yi;theta) / F(k;theta) ]???
Second question, concerning model estimation with md and weight
function, C15, p.375
F(x);theta)=1-e^(-x / theta)
estimate theta by md:
w1[Fn(500)-F(500;theta)]^2+w2[Fn(2000)-F(2000;theta)]^2
the weights w1 and w2 are inversly proportional to the variance and
theta=1000 find the weights w1 and w2
you get
w1 inversly prop. to 4.1902n
w2 inversly prop. to 8.5455737n
the part i don't get is after this: shouldn't n be 2? why do they do
the following thing
w1=4.1901 / (4.1901+8.5474)=0.33
w2=8.5474 / (4.1901+8.5474)=0.67
Thanks for help
Dimitra
dr@g-g-a.com