Multi-Scale Data enhancement through SIAC-MRA

：Jennifer K. Ryan（瑞典皇家理工学院）
：2024-03-11 16:30
：海韵园实验楼105报告厅

报告人：Jennifer K. Ryan（瑞典皇家理工学院）

时  间：2024年3月11日16:30

地  点：海韵园实验楼105报告厅

内容摘要:

Extracting extra information from data can allow for more insight into the interaction between disparate scales. It can also aid in minimising error and decreasing noise in data. While the ability to move data from fine resolutions to coarser resolutions is straight forward utilizing a multi-resolution analysis framework, moving data from a coarse resolution to a finer resolution while reducing errors is more challenging. This relies on combining filtering techniques into the multi-resolution analysis framework [Ryan, CAMC 2022; Picklo & Ryan, SISC 2022]. This approach has the further advantage of requiring fewer computations to gain insight into calculations such as for Bohm speed [Picklo et al. JCP 2024]. In this talk, we present methods for data enhancement through multi-resolution analysis and the Smoothness-Increasing Accuracy-Conserving (SIAC) filtering framework. SIAC is known toinherently take advantage of the underlying physics and allow for the full resolution of the approximation and its derivatives in both the physical domain and Fourier signal space. We discuss recent advances and reliance of the approach on the underlying numerical method that generated the data. 

个人简介：

Prof. Jennifer K. Ryan serves as a Professor and Division Head in Numerical Analysis, Optimization and Systems Theory of the Department of Mathematics at KTH Royal Institute of Technology. She obtained her Ph.D. sin Applied Mathematics from Brown University in 2003. Prior to that, Prof. Ryan an M.Sc. and a B.Sc. at Rutgers University and New York University respectively. Prior to joining KTH, she worked as a Householder Fellow at Oak Ridge National Laboratory, a Assistant Professor at Virginia Polytechnic Institute & State University, a Assistant Professor at Delft University of Technology, a Senior Lecturer at University of East Anglia, and a Associate Professor at Colorado School of Mines. Prof. Ryan's research concentrates on designing and developing numerical schemes for various applications,  the majority of which is focused on extracting extra accuracy out of data obtained from numerical simulations. She also investigates the underlying theory, not only to improve the method itself, but also to improve its computational efficiency. She does this by using the inherent properties of the method. The techniques that she develop can be exploited for use in imaging, data extraction, detection of discontinuous phenomenon, visualisation of fluids, and more. She has published more than 40 articles in top-ranked international journals, including JCP, SINUM, SISC, MC, etc.

 

联系人：赵状