FREE WEBINAR: Machine Learning for Narrowing the Simulation-Test Gap in Digital Twins

Learn the basics of statistical model calibration to quantify uncertainties in simulation.

This webinar will provide an introduction to the basics of statistical calibration. Through the use of several example problems, the underpinning ideas and benefits of statistical calibration will be illustrated. This includes the unique ability of statistical calibration, as compared to alternative calibration approaches, to account for model form uncertainty in addition to parameter uncertainty.

SmartUQ presents using machine learning to make predictive digital twins more accurate.

DATE: August 24, 2021
TIME: 03:00 PM EDT/ Noon PDT

Establishing how well a numerical simulation represents reality is critical for making simulation results more trustworthy for decision makers.

This webinar will focus on statistical model calibration, a machine learning process used to quantify uncertainties (both parameter and model form) in simulations, a means to narrow the gap between simulation and physical test outcomes.

This process works not just for making simulations more accurate for traditional uses such as in the design phase, but also for making models more accurate when used as part of a digital twin workflow. Calibrated predictive models of a simulation can even themselves be used as digital twins, allowing for rapid predictions.

Using examples, this LIVE webinar with Q&A covers the underpinning ideas and benefits of statistical calibration such as:

  • Unique ability to account for both parameter and model form uncertainties.
  • Frequentist and Bayesian calibration options (available in SmartUQ)
  • Applications of statistical calibration to digital twins.

Share This Article

Subscribe to our FREE magazine, FREE email newsletters or both!

Join over 90,000 engineering professionals who get fresh engineering news as soon as it is published.