How do I compute P(y>x) given f(x) and f(y|x)?

Click For Summary

Homework Help Overview

The problem involves computing the probability P(y>x) given the probability density functions f(x) = e^-x and f(y|x) = 1/x e^(-y/x). The discussion includes three parts: computing the joint density of (x,y), finding the expected value E(y), and specifically addressing the computation of P(y>x).

Discussion Character

  • Exploratory, Assumption checking, Problem interpretation

Approaches and Questions Raised

  • Participants discuss the need to derive f(y) to compute P(y>x) and question whether it is possible to find P(y>x) without this derivation. There are attempts to explore the relationship between f(x) and the intersection of functions to aid in understanding the problem.

Discussion Status

The discussion is ongoing, with participants providing hints and exploring different interpretations of the problem. There is no explicit consensus on the approach to take for computing P(y>x), and various lines of reasoning are being considered.

Contextual Notes

Participants are navigating the constraints of the problem, including the need for specific probability distributions and the implications of the given functions. There is uncertainty about the necessary steps to derive the required probability.

BookMark440
Messages
9
Reaction score
0

Homework Statement


Given f(x) = e^-x and f(y|x) = 1/x e^(-y/x). Three parts: (a) Compute density of (x,y), (b) Compute E(y) and (c) Compute P(y>x).


Homework Equations


f(x,y) = f(y|x)f(x)
if f(x) = ve^(-vx), then E(x)=v^(-1)

The Attempt at a Solution



I'm stuck on a problem. I was given f(x) and f(y|x) and was able to derive f(x,y) and compute E(y). The third step of the problem is computing P[y>x]. I think I need to know f(y) to answer this problem but I can't figure out how to derive it. Or is there a way to compute P(y>x) given the info I know without deriving f(y)?
 
Physics news on Phys.org
Hi BookMark440! :smile:

Hint: what is P(y>x) for a fixed value of x? :wink:
 
Maybe I am looking for the wrong answer. Do I simply need to find the portion of f(x)=exp(-x) that are to the left of its intersection with f(x)=x ?

This would mean I am looking for the intersection, which is exp(-x) = x, solving for x?

That is:

-x*log e = log x
0.4343 = -(log x)/x

But what step is next? Is there some law of logs I am missing to simplify this? From graphing the problem, it looks like the point of intersection is (1 - log e) but I can't see how the above translate into (1-log e).
 
Last edited:
BookMark440 said:
Maybe I am looking for the wrong answer. Do I simply need to find the portion of f(x)=exp(-x) that are to the left of its intersection with f(x)=x ?

No, for P(y>x) you need to find the portion of y, don't you? :smile:
 

Similar threads

  • · Replies 6 ·
Replies
6
Views
2K
Replies
8
Views
2K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 8 ·
Replies
8
Views
1K
  • · Replies 18 ·
Replies
18
Views
3K
  • · Replies 8 ·
Replies
8
Views
2K
  • · Replies 4 ·
Replies
4
Views
1K
Replies
2
Views
1K
Replies
9
Views
2K
Replies
3
Views
1K