Probability generating functions
Maths and statistics discussion, revision, exam and homework help.
-
Probability generating functions
This isn't really the original Achilles & tortoise problem but it uses the characters:
Suppose that the race began with Achilles waiting on the starting line while the tortoise was given a start of t time units, where t is a positive integer. Suppose also that, during each time unit, the tortoise either moves 1 metre with probability p or stays where it is with probability 1-p. Write down the distribution of the distance travelled by the tortoise before it fails to move and find the probability generation function.
I could do this part because it's the geometric distribution, and the pgf is (1-p)/(1-sp).
Once he has started, Achilles runs at a constant rate of 1m per time unit. Find the pgf for the distance from the start at which Achilles catches the tortoise and show that its expectation is tp/(1-p).
I'm really stuck on this part.. Suppose the tortoise travels D metres before Achilles starts. Achilles will need to travel each time the tortoise travels, plus all the times the tortoise fails to travel, and there need to be D of these, so we need to find the amount of time it takes the tortoise to fail D times? It looks like a sum of I.I.D. random variables but I'm not sure how to go about it. Please help? -
Re: Probability generating functions
Basically, every time the tortoise fails to move, Achilles catches up one unit. And you know he needs to catch up t units. So it is exactly a sum of IID random variables (of the form you were asked about in the first part of the question).
What exactly, are you stuck on? -
Re: Probability generating functionsWhy does he need to catch up t units? The tortoise has not necessarily moved t units in the headstart, so the number of units he needs to catch up on looks like a random variable as well?(Original post by DFranklin)
Basically, every time the tortoise fails to move, Achilles catches up one unit. And you know he needs to catch up t units. So it is exactly a sum of IID random variables (of the form you were asked about in the first part of the question).
What exactly, are you stuck on? -
Re: Probability generating functionsBut each time unit corresponds to Achilles catching up 1 distance unit. He has a headstart of D~Bi(t, p) units, of Achilles needs to catch up on D time (or, equivalently, distance) units? I don't think he needs to catch up t units because the tortoise is not necessarily that far ahead. Maybe there's something I'm missing?(Original post by DFranklin)
I haven't done the calcs, so this is possibly wrong, but think in terms of catching up *time* units, rather than distance.
[That is, each time the tortoise fails to move, he's "wasting" a unit of time, relative to Achilles].