BGonline.org Forums

RANDOM NUMBER GENERATION

Posted By: Timothy Chow
Date: Tuesday, 22 June 2010, at 5:24 p.m.

In Response To: RANDOM NUMBER GENERATION (Fabrice Liardet)

Kolmogorov complexity is certainly one way to approach the definition of randomness. But as you say, it is uncomputable, so it is not of much use in the practical design of random number generators. In practice we don't really care about "patterns" in the random string that cannot be detected in a feasible amount of time. Therefore, immunity to all polynomial-time statistical tests is the usual criterion for randomness used by, say, cryptographers. This is basically the computational-complexity analogue of Kolmogorov complexity.

Messages In This Thread

• RANDOM NUMBER GENERATION
  Ray Kershaw -- Monday, 21 June 2010, at 7:14 a.m.
  • RANDOM NUMBER GENERATION
    Maik Stiebler -- Monday, 21 June 2010, at 12:51 p.m.
  • RANDOM NUMBER GENERATION
    Sam Pottle -- Monday, 21 June 2010, at 4:37 p.m.
  • RANDOM NUMBER GENERATION
    Jason Lee -- Monday, 21 June 2010, at 4:56 p.m.
  • RANDOM NUMBER GENERATION
    Frank Berger -- Monday, 21 June 2010, at 9:48 p.m.
  • RANDOM NUMBER GENERATION
    Timothy Chow -- Monday, 21 June 2010, at 11:46 p.m.
  • RANDOM NUMBER GENERATION
    Fabrice Liardet -- Tuesday, 22 June 2010, at 11:51 a.m.
    • RANDOM NUMBER GENERATION
      Timothy Chow -- Tuesday, 22 June 2010, at 5:24 p.m.
      • RANDOM NUMBER GENERATION
        Fabrice Liardet -- Tuesday, 22 June 2010, at 7:31 p.m.