Short Course: Uncertainty Quantification and Inverse problems
Location: A817, School of Aeronautics
Time: 27th, Nov, 9:00-12:00 14:30-17:30
Reporter: Richard P. Dwight from TU Delft
Invited by: Prof. Zhonghua Han
Sponsored by: The overseas Expertise Introduction Center for Discipline Innovation on Complex Flow and Its Control (the 111 Center), School of Aeronautics
Abstract: This course is a short introduction to the field of uncertainty quantification, with emphasis on techniques for model calibration and large-scale inverse problems with the Bayesian framework. Lectures are complemented by 3 tutorial exercises in Python, which exploit the theory learned. Prerequisites are: (1) a first-course on probability/statistics (though this material will be briefly refreshed in the first lecture), (2) experience with Python (and numpy) programming would be very beneficial for the tutorials. If 1) is missing I recommend reading Chapter 4 of the textbook Ralph C. Smith "Uncertainty Quantification" in advance of the course.
Schedule:
|
Time |
Arrangement |
27th, Nov |
9:00-10:00 |
Lecture 1: Introduction to UQ and its relevance in aerospace |
10:30-12:00 |
Lecture 2: Bayesian approaches to regression problems |
14:30-17:30 |
Lecture 3: Large-scale stochastic inverse problems with high-dimensional randomness (and random fields) |
|
|
Introduction of Prof. Dwight:
Professor Richard P. Dwight received his Ph.D. from the University of Manchester in 2006, worked as an associate professor in the Department of aerodynamics at the school of Aerospace Engineering, Delft University of technology since 2009, and as a visiting professor at the Centrum voor Wiskunde en Informatica (CWI) in The Netherlands since 2017. Professor Richard P. Dwight has made great achievements in computational fluid dynamics, machine learning, aerodynamic design optimization, surrogate-based optimization method, adjoint method, uncertain optimization method and other fields.
Contact:
Name: Fei Liu
Phone: 18729537605
Wechat: ltwbtf
