ABSTRACT
A simulation can be deterministic, stochastic, or mixed. Stochastic modeling recognizes the inherent uncertainty in parameters and relationships. For many modeling purposes, the stochastic nature of a system is an essential feature to be represented. This chapter discusses the following concepts: randomness as the underlying construct for probability distributions, basic concepts of probability distributions with some common probability distributions, and use of probability distributions in Monte Carlo trials. A uniform distribution represents a range of equally likely outcomes with an equal number of measures in each interval. The symmetric normal distribution is used for outcomes likely to occur on either side of the average value. The lognormal distribution is appropriate for modeling parameters with long right tails in their distributions. The PERT probability distribution, which is used in many modeling and project scheduling tools, is a simplified form of a Beta distribution.
