Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Convergence P-a-s

  1. May 18, 2012 #1
    So I have a definition;
    Xn n=1,2.... is a sequence of random variables on ( Ω,F,P) a probability space, and let X be another random variable.
    We say Xn converges to X almost surely (P-a-s) iff P({limn →∞ Xn=X}C) = 0

    It then goes on to say that checking this is the same as checking
    limm →∞ P({Supn≥m|Xn-X| ≥ε }) = 0
    Can somebody please explain why this is true, I don't understand at all how to get from one to the other properly.

  2. jcsd
  3. May 18, 2012 #2
    They're just two ways to express the same concept; that you can get Xn as close to X as you want with an n sufficiently large.
  4. May 19, 2012 #3

    Stephen Tashi

    User Avatar
    Science Advisor

    "It" isn't being very precise. I suppose mathematical tradition tells us that [itex] \epsilon [/itex] is a number greater than zero. Tradition also tells us that the quantifier associated with [itex] \epsilon [/itex] is "for each", so that's a hint about what it means. The notation it is using for sets is very abbreviated. The usual notation would tell us that a set is "the set of all.... such that ....".

    If you want to understand the assertions precisely, the first thing you must do is to understand precisely what they assert. I don't know if that is your goal. If it is, try to write out exactly what each of those statements claims using better notation.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook