Experimental Design and Simulation Metamodeling

This portion of the class focuses on modeling and analysis of systems using discrete event simulation techniques. Topics that are covered include input data modeling, random number and random variable generation, output analysis, variance reduction techniques, and experimental design. One of the topics where machine-learning techniques can be incorporated into this class is experimental design and regression analysis. Design of experiments can be used to identify what factors have a significant impact on the performance measure of the interest. Moreover, metamodeling techniques such as regression analysis and neural networks can be used to develop a mathematical relationship between certain inputs and outputs of the simulation models. Some of the topics mentioned above are included in the files below.

|Lecture 1|

|Lecture 2a|

|Lecture 2b|

|Simulation Homework|