Actuarial Outpost Conditional PDFs
 Register Blogs Wiki FAQ Calendar Search Today's Posts Mark Forums Read
 FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions

 DW Simpson & CoWorldwide ActuarialRecruitment Entry Level JobsCasualty, Health, Life, Pension,Investment --Insurance / Consulting Asian JobsHong Kong, China, India, Japan, Korea, Indonesia, Singapore,Malaysia, and more Registration FormBe Notified of New Actuarial Jobs

#1
09-16-2009, 03:40 AM
 Sandimashischool Member Join Date: Feb 2007 Posts: 547
Conditional PDFs

Let Y be a gamma random variable with parameters (s, a). That is, its density is:
f(y) = C*exp(-ay)y^(s-1), y>0

where C is a constant that does not depend on y. Suppose also the conditional distribution of X given Y = y is Poisson with mean y. That is,

P[X=x|Y=y] = (exp(-y)*y^x)/x! x >=0

Show that the conditional distribution of Y given that X = x is the gamma distribution with parameters (s+x, a+1)

----------------------------------
Ok so here is what I figured out,

I need f(y|x) = f(x,y)/f(x) = (f(x|y) * f(y))/f(x)

I know f(x|y) = (exp(-y)*y^x)/x!
f(y) = C*exp(-ay)y^(s-1) ~and if its gamma then C = (a^s)/(s-1)!

So when I multiply f(x|y) and f(y), and move constants to the left I get

(a^s)/((s-1)!*X!) * exp(-y(1+a))*y^(s-1) ..... and now I am stuck

Is this my final f(x,y), and even if it is I need to get f(x) to divide so I need to integrate 0 to Infinity to get the marginal distribution (I think) but I don't know how to integrate that (by parts maybe? ... I'm guessing I made a mistake somewhere because I can't integrate that) Any advice?

PS thanks to all those geniuses helping me out there (jraven, David, Actuarialsuck, etc) I really appreciate it.
#2
09-16-2009, 09:45 AM
 lttownsend Member SOA ASPPA Join Date: Aug 2009 Location: Chicago Favorite beer: Gin Posts: 325

The formula for conditional density f(y|x) = f(x,y)/f(x) only holds for continuous distributions, I think. Since X is discrete, that would seem to throw a wrench in this formulation.
#3
09-16-2009, 10:17 PM
 TwoStep Member CAS SOA Join Date: Oct 2008 Studying for All done! Posts: 793

X is a gamma/poisson mixture. I remember this from exam C, but I didn't realize it was on P. It turns out the unconditional distribution of X is negative binomial:

f(x) = p^s*(1-p)^x*(s+x-1)!/(x!*(s-1)!) where p = 1/(a+1)

It's been a while, I don't really know where to go from here, but I hope that helps. Also, double check this link to make sure I didn't make an error in the formula:

http://en.wikipedia.org/wiki/Negativ...l_distribution

 Thread Tools Display Modes Linear Mode

 Posting Rules You may not post new threads You may not post replies You may not post attachments You may not edit your posts BB code is On Smilies are On [IMG] code is On HTML code is Off

All times are GMT -4. The time now is 10:13 AM.

 -- Default Style - Fluid Width ---- Default Style - Fixed Width ---- Old Default Style ---- Easy on the eyes ---- Smooth Darkness ---- Chestnut ---- Apple-ish Style ---- If Apples were blue ---- If Apples were green ---- If Apples were purple ---- Halloween 2007 ---- B&W ---- Halloween ---- AO Christmas Theme ---- Turkey Day Theme ---- AO 2007 beta ---- 4th Of July Contact Us - Actuarial Outpost - Archive - Privacy Statement - Top