Department Seminar: Qiang Zhou, Univ Of Wisconsin - Madison - Monday, Feb 14 @ 8:00am
Surrogate Modeling for Computer Experiments with Quantitative and Qualitative Input Variables
Abstract: Driven by rapid development of computing and mathematical techniques, computer simulations have emerged as a critical tool in engineering design. To cut down simulation times, surrogate models are widely used for building approximations of time-consuming computer simulation codes in a large array of engineering design problems. The prevailing framework for building surrogate models deals almost exclusively with quantitative design factors only. However, many engineering design problems involve both quantitative and qualitative design factors. In this talk, we will introduce an easy-to-implement surrogate modeling method for computer experiments with both quantitative and qualitative input variables. The method can also be adopted as an integrated meta-modeling tool for computer simulations with multivariate output to improve surrogate model prediction accuracy. Novel experimental design and model selection procedures for such surrogate modeling scheme will also be discussed. Case studies have been conducted to demonstrate the effectiveness of our methods. Some other projects in quality engineering will also be briefly discussed in the end of this talk.
Audience-Based Site-Wide Navigation:back to top