Department of Industrial Engineering,Sharif University of Technology
A known theorem in probability is adopted and through a probabilistic approach, it is generalized
to develop a method for generating random deviates from the distribution of any continuous
random variable. This method, which may be considered as an approximate version of the
Inverse Transform algorithm, takes two random numbers to generate a random deviate, while
maintaining all the other advantages of the Inverse Transform method, such as the possibility of
generating ordered as well as correlated deviates and being applicable to all density functions,
regardless of their parameter values.