2
$\begingroup$

I would like to check if my calculations are correct: We have two buses, arrivals of bus (A) form a Poisson process of rate $1$ bus per hour and arrivals of bus (B) form an indep. Poisson process of rate $7$ buses per hour.

(1) Probability that exactly 3 buses of type (B) pass while I still wait for bus (A)

Well bus (B) has rate 7, bus (A) has rate 1, therefore $(\frac{7}{7+1})^3$

(2) Probability that exactly 5 buses of typ (A) pass in 1 hour.

My approach was to calculate $P(B=k|A+B=5)$ for $k\in[0,5]$ and add everything together, so I get $\binom 5 0 (\frac{7}{7+1})^0(\frac{7}{7+1})^5+...+\binom 5 5(\frac{7}{7+1})^5(\frac{7}{7+1})^0$, correct?

$\endgroup$
  • $\begingroup$ For 1) you have to calculate the expected waiting time for a bus A. It is longer then 1/2 an hour as say by chance bus A comes after 10 min, then again after 1 hour 50 min. You are more likly to arrive at the bus stop in the 1 hour 50 min gap. In 2) nothing depends on bus B so it is P(k<6) for bus A $\endgroup$ – user121049 Feb 10 '14 at 12:26
  • $\begingroup$ Thanks for your comment, would you mind to write it down formally as an answer? So I can give you the bounty. $\endgroup$ – TI Jones Feb 10 '14 at 14:10
  • $\begingroup$ ;) I'll try and find the answer for the expected waiting time. $\endgroup$ – user121049 Feb 10 '14 at 14:41
  • $\begingroup$ The expected waiting time for bus A is 1 hour given a Poisson distribution. I.e twice what one may have naively assumed. $\endgroup$ – user121049 Feb 11 '14 at 19:47
2
+50
$\begingroup$

For 1) you have to calculate the expected waiting time for a bus A. It is longer then 1/2 an hour as say by chance bus A comes after 10 min, then again after 1 hour 50 min. You are more likly to arrive at the bus stop in the 1 hour 50 min gap. In 2) nothing depends on bus B so it is P(k<6) for bus A.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ can u explain bit more detail, I want to learn it. thank you $\endgroup$ – STEEL Feb 17 '14 at 4:14
  • $\begingroup$ If you have a Poisson process then the interval arrival times are from an exponential distribution. You can of course have more general distribution with mean $\mu$ and variance $\sigma^2$ in which case the expected waiting time for a bus (given that you arrive at random) is given by $\frac{1}{2}\mu(1+(\frac{\sigma}{\mu})^2)$. So the greater the variance the longer you have to wait. This is because you are more likely to arrive in a big gap between buses. The maths is the same as for the "Inspection Paradox" of which google will tell you more. $\endgroup$ – user121049 Feb 17 '14 at 15:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.