3.9 Summary

In this chapter, learned how to use KSL constructs to collect statistics within Monte Carlo models. We use the KSL’s random variate generation methods and the functionality for collecting and reporting statistics to easily setup and perform Monte Carlo experiments. The quantities to be estimated from the Monte Carlo experiments are random variables. Thus, the Monte Carlo procedure generates a random sample of from the experiment. The random sample from the experiment provides estimates such as the sample average and other statistical quantities. For example, to estimate a probability, we used an indicator variable and computed statistics on the observations of the indicator variable.

When executing a Monte Carlo experiment, we should report the sampling error or a confidence interval on the quantities estimated from the Monte Carlo experiment. In addition, because we are performing an experiment, we should plan the sampling by understanding the sampling error and if possible, pre-plan the required sample size. Besides the random variate generation and statistical classes within the KSL, we can use the KSL’s MCExperiment class to execute Monte Carlo experiments.

In addition, we explored how to develop models in for which time is not a significant factor. In the case of the news vendor problem, where we simulated each day’s demand, time advanced at regular intervals. In the case of the area estimation problem, time was not a factor in the simulation. These types of simulation experiments are often termed static. In the next chapter, we begin our exploration of simulation experiments where time is an integral component in driving the behavior of the system. In addition, we will see that time will not necessarily advance at regular intervals (e.g. hour 1, hour 2, etc.). This will be the focus of the rest of the book.