Department of Industrial Engineering
4207 Bell Engineering Center
1 University of Arkansas
Fayetteville, AR 72701
Phone: (479) 575-3156
Fax: (479) 575-8431
PhD Qualifying Exam Information for Simulation
Materials Permitted:
- Hard copy (typed or hand written) notes covering the topics below and created by the student taking the exam.
- Laptop computer for model building use only
- Calculator
1. Modeling Randomness
- Be able to describe what a random number seed is and how to generate random numbers from different streams.
- Given a probability distribution, be able to specify an algorithm for generating random variates from the distribution via the inverse transform method, convolution, mixtures, truncated, shifted, acceptance/rejection methods as deemed appropriate.
- Be able to describe the construction of P-P, and Q-Q Plots and be able to interpret their meaning within the context of input modeling
- Be able to perform an input distribution fitting exercise and make an input model recommendation, justified by statistical analysis
2. Statistical Analysis
- Be able to compute and interpret the meaning of confidence intervals
- Be able to compute the sample size necessary to estimate a desired output statistic to within +/- desired half-width.
- Be able to determine the overall confidence level associated with making a decision based on a set of confidence intervals
- Be able to set the individual confidence interval levels in order to ensure an overall confidence level on a decision based on a set of confidence intervals
- Be able to perform and interpret a basic multiple comparison analysis
- Be able to describe, discuss, and interpret a Welch plot to identify a possible warm up period for an infinite horizon steady state simulation
- Be able to set up an experimental design to perform a sensitivity analysis of a simulation according to desirable DOE criteria.
- Be able to interpret the results of a main effect experimental design analysis of a simulation.
3. Be able to build a model of a static (Monte Carlo) simulation situation and perform a statistical analysis of the results.
4. Be able to model discrete event dynamic systems commonly found within industrial engineering
- Be able to describe a system in terms of its transient, structural, attributes, resources, activities, events, processes, state variables, etc.
- Be able to represent the system within an object-oriented event-scheduling paradigm
- Be able to verify and validate a DEDS model
- Be able to exercise and interpret the results of a DEDS