BEGIN:VCALENDAR
PRODID:-//eluceo/ical//2.0/EN
VERSION:2.0
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:4fcd06c52d53b3f280c63ede24cafb44
DTSTAMP:20240302T201625Z
SUMMARY:Nelder Lecture Series – Prof Nathan Kutz: Data-driven modeling of
dynamical systems
DESCRIPTION:Professor Nathan Kutz will hold series of four lectures on Tues
days from 14 November until 5 December 2023.\nLecture Room 402 in the CDT
space:\nEnter Sherfield building from the walkway\, on the other side of t
he staircase on the righthand side you see the door which you need to ente
r\, take a stairs or a lift to the 4th floor and you see the door to the C
DT space. You look for the room there.\nAbstract: Data-driven models a
re critically enabling in many application areas where the underlying dyna
mics are unknown or only partially known\, or where high-fidelity simulati
ons are computationally expensive to generate. The ability to produce ac
curate\, low-rank\, proxy models enable dynamic models to transform the re
presentation and characterization of such systems. Data-driven algorithm
s have emerged as a viable and critically enabling methodology that is typ
ically empowered by machine learning algorithms. Indeed\, there are a di
versity of mathematical algorithms that can be used to produce data-driven
models including (i) dynamic mode decomposition\, (ii) sparse identificat
ion for nonlinear dynamics\, and (iii) neural networks. Each of these m
ethods are highlighted here with a view towards producing proxy\, or reduc
ed order\, models that enable efficient computations of high-dimensional s
ystems. Moreover\, these methods can be used with direct measurement dat
a\, computational data\, or both in generating stable representations of t
he dynamics. The course will focus on theory and computation with imple
mentation of algorithms in python using a combination of packages (pyDMD\,
pySINDy) and deep learning algorithms (PyTorch)\nBio: Nathan Kutz is th
e Yasuko Endo and Robert Bolles Professor of Applied Mathematics and Elect
rical and Computer Engineering at the University of Washington\, having se
rved as chair of applied mathematics from 2007-2015. He is also the Direct
or of the AI Institute in Dynamic Systems (dynamicsAI.org). He received
the BS degree in physics and mathematics from the University of Washington
in 1990 and the Phd in applied mathematics from Northwestern University i
n 1994. He was a postdoc in the applied and computational mathematics prog
ram at Princeton University before taking his faculty position. He has a w
ide range of interests\, including neuroscience to fluid dynamics where he
integrates machine learning with dynamical systems and control. During th
e academic year 2023-2024\, he will be a visiting professor at Imperial Co
llege London and the Alan Turning Institute.\n
URL:https://www.imperial.ac.uk/events/169392/nelder-lecture-series-prof-nat
han-kutz-data-driven-modeling-of-dynamical-systems-2/
DTSTART;TZID=Europe/London:20231121T100000
DTEND;TZID=Europe/London:20231121T120000
LOCATION:Lecture Room 402 in the CDT space \, Sherfield Building\, South Ke
nsington Campus\, Imperial College London\, London\, SW7 2AZ\, United King
dom
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:STANDARD
DTSTART:20231121T100000
TZNAME:GMT
TZOFFSETTO:+0000
TZOFFSETFROM:+0000
END:STANDARD
END:VTIMEZONE
END:VCALENDAR