For many applications in science and engineering a system’s response may feature large discontinuities. For example, when the failure mode of a component changes, there is likely to be a discontinuity in the failure stress. The switching of vibrational modes can also cause discontinuities in the response. However, many common machine learning (ML) modeling techniques such as Gaussian process (GP) modeling struggle with capturing the behavior of discontinuous response profiles.
SmartUQ has developed a unique GP based model, the Mixed Input Classification Emulator (An emulator is a predictive model), specifically designed for problems with discrete response behavior such as discontinuities. Join us for this webinar to learn more and see a demo.
