Monte Carlo Method
The drawback to using Equation
(A.2) to determine generating function for a particular distribution
is that solving for the random variable x in terms of R is difficult.
A second method, similar to the way integration is done stochastically,
is to generate pairs of random numbers. A point R1 is generated
such that R1 is uniformly distributed over the interval
(a,b) and R2 over the interval from 0 to the maximum value that
the function p takes on in (a,b). If
then then R1 is the new random variable. Otherwise new points
R1 and R2 are generated until the condition (A.14) is
satisfied. The speed of this technique depends on how many
unsatisfactory points R1 and R2 must be discarded until a good
pair is obtained [6].
|