Award
2020 NSF CAREER Award Winner
Department
Grado Department of Industrial and Systems Engineering
Awarded Project
Model-Free Input Screening and Sensitivity Analysis in Simulation Metamodeling
Simulation is a widely used method to model stochastic systems, but according to Xi Chen, analysis of these models becomes computationally and statistically difficult when the models involve a large number of potential input parameters. In order to improve model tractability, Chen believes it is important to identify a subset of significant input parameters and to use them to design effective simulation experiments.
With her NSF CAREER award, Chen will work to provide new input screening and sensitivity analysis techniques for improving the decision-making capability within performance and schedule requirements in large-scale, complex systems applications. Long-term, Chen believes her research can contribute to national health and prosperity by improving modeling and analysis using large-scale simulations.
What path did you take to get to this point in your career and research?
Growing up, I have always been fascinated by the ability of mathematics to concisely describe relationships and derive useful, generalized conclusions based on simplified assumptions and rigorous arguments. My Ph.D. study in industrial engineering and management sciences at Northwestern University really opened my eyes to how operations research, as a subfield of applied mathematics, could be used to make significant impacts in the real world. And I have enjoyed my job as a professor to continue exploring this field and make my unique contributions.
What impact do you hope your research will have?
With recent breakthroughs in high-performance computing, great strides have been made in successfully investigating computationally-intensive problems in a wide range of scientific disciplines. The idea of using a “digital twin,” a simulation model of a physical system that represents all of its functionalities, is permeating almost everywhere in the industrial and public sectors, for simulating real-world conditions to analyze system responses to changes and improve operations. Yet, our ability to use a simulation model to understand behaviors of a complex stochastic system in a virtually interactive way is not ready, due to the following challenges: an enormously large number of input combinations to evaluate; a high computational cost associated with executing a single simulation run; and the stochastic nature of the simulation model.
My research will provide new input screening and sensitivity analysis techniques for improving our decision-making capability to meet challenging performance and schedule requirements in real-world applications. The techniques will apply to a wide range of application areas, such as biomedical studies, health care, manufacturing, and defense and homeland security operations.
What do you find most interesting about the field of engineering in which you study?
I like my field, operations research, because it allows us to use advanced analytical methods to help make better decisions. My specific area, stochastic simulation methodology and analysis, allows me to use mathematics to answer questions about how to collect robust data intelligently and make more effective, actionable decisions based on fuller consideration of available options in a timely fashion.
What's a common myth or misconception about the subject of your research that you'd like to debunk?
People often have the misconception that computer programming or coding plays the major role in simulation-based modeling and analysis. In fact, having a solid background in mathematics and statistics proves more useful for pursuing your interest in this field, whether it lies in particular applications or underlying theory and methodology.