BUILDING CREDIBLE INPUT MODELS

Most discrete-event simulation models have stochastic elements that mimic the probabilistic nature of the system under consideration. A close match between the input model and the true underlying probabilistic mechanism associated with the system is required for successful input modeling. The general question considered here is how to model an element (e.g., arrival process, service times) in a discreteevent simulation given a data set collected on the element of interest. For brevity, it is assumed that data is available on the aspect of the simulation of interest. It is also assumed that raw data is available, as opposed to censored data, grouped data, or summary statistics. This example-driven tutorial examines introductory techniques for input modeling. Most simulation texts (e.g., Law and Kelton 2000, Fishman 2001) have a broader treatment of input modeling than presented here. Nelson and Yamnitsky (1998) survey advanced techniques. 1 DATA COLLECTION There are two approaches that arise with respect to the collection of data. The first is the classical approach, where a designed experiment is conducted to collect the data. The second is the exploratory approach, where questions are addressed by means of existing data that the modeler had no hand in collecting. The first approach is generally better in terms of control and the second approach is generally better in terms of cost. Collecting data on the appropriate elements of the system of interest is one of the initial and pivotal steps in successful input modeling. An inexperienced modeler, for example, collects wait times on a single-server queue when waiting time is the measure of performance of interest. Although these wait times are valuable for model validation, they do not contribute to the input model. The appropriate data elements to collect for an input model for a singleserver queue are typically arrival and service times. A