Introduction to Uncertainty Quantification Techniques
This mini-tutorial will focus on concepts, methods, and applications in the important and rapidly growing field of uncertainty quantification (UQ) techniques for numerical computation. The presentations will be introductory in nature, appropriate for those new to the area as well as those seeking a high level overview of various techniques that are used today, or are currently under development. Since this is an introductory overview, the emphasis in this tutorial will be on practical ideas.
Topics will include:
- An introduction to UQ concepts, including aleatory and epistemic ideas
- Classical UQ sampling and reliability methods such as Monte Carlo, Latin Hypercube, and Mean Value
- Stochastic Galerkin and stochastic collocation methods
- Bayesian methods
- A brief introduction to alternative non-probabilistic approaches and current research areas
Attendees will gain an understanding of how these various UQ methodologies are formulated and applied to problems in computational science and engineering.
Organizer: Gianluca Iaccarino, Stanford University
Alireza Doostan, Stanford University
Michael S. Eldred, Sandia National Laboratories
Omar Ghattas, University of Texas at Austin
Gianluca Iaccarino, Stanford University