BEGIN:VCALENDAR
VERSION:2.0
PRODID:researchseminars.org
CALSCALE:GREGORIAN
X-WR-CALNAME:researchseminars.org
BEGIN:VEVENT
SUMMARY:Johan Ulander (Chalmers University of Technology)
DTSTART:20230830T111500Z
DTEND:20230830T120000Z
DTSTAMP:20260416T215336Z
UID:cam/1
DESCRIPTION:Title: Boun
dary-preserving numerical schemes for stochastic (partial) differential eq
uations\nby Johan Ulander (Chalmers University of Technology) as part
of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nIn this talk we con
sider stochastic (partial) differential equations whose solutions remain i
n some (half-)bounded domain. This includes\, for example\, models for hea
t flow whose solutions remain positive. In general\, classical numerical s
chemes do not have the property of remaining in such domain. In this half-
way seminar\, I present some novel ideas for developing and analysing boun
dary-preserving numerical schemes for stochastic (partial) differential eq
uations. The presentation is based on joint works with Charles-Edouard Br
éhier\, David Cohen\, and Lluís Quer-Sardanyons.\n
LOCATION:https://researchseminars.org/talk/cam/1/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Yoshio Komori (Kyushu Institute of Technology)
DTSTART:20230906T111500Z
DTEND:20230906T120000Z
DTSTAMP:20260416T215336Z
UID:cam/2
DESCRIPTION:Title: Spli
t S-ROCK methods for high-dimensional stochastic differential equations\nby Yoshio Komori (Kyushu Institute of Technology) as part of CAM semina
r\n\nLecture held in MV:L14.\n\nAbstract\nWe propose explicit stochastic R
unge--Kutta (RK) methods for high-dimensional It\\^{o} stochastic differen
tial equations. By providing a linear error analysis and utilizing a Stran
g splitting-type approach\, we construct them on the basis of orthogonal R
unge--Kutta—Chebyshev methods of order 2. Our methods are of weak order
2 and have high computational accuracy for relatively large time-step size
\, as well as good stability properties. In addition\, we take stochastic
exponential RK methods of weak order 2 as competitors. It is shown that th
e proposed methods can be very effective on high-dimensional problems whos
e drift term has eigenvalues lying near the negative real axis and whose d
iffusion term does not have very large noise. This is a joint work with Pr
of. Kevin Burrage.\n
LOCATION:https://researchseminars.org/talk/cam/2/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Sagy Ephrati (University of Twente)
DTSTART:20230920T111500Z
DTEND:20230920T120000Z
DTSTAMP:20260416T215336Z
UID:cam/3
DESCRIPTION:Title: Stoc
hastic modeling for coarse computational geophysical fluid dynamics\nb
y Sagy Ephrati (University of Twente) as part of CAM seminar\n\nLecture he
ld in MV:L14.\n\nAbstract\nStochasticity has been employed systematically
in geophysical fluid dynamics (GFD) to model uncertainty. Additionally\, f
ully resolving geophysical flows is computationally expensive due to the l
arge range of scales of motion present in these flows. These computational
costs are efficiently mitigated by performing GFD simulations on coarse c
omputational grids and modeling the effects of unresolved scales on resolv
ed scales. On such grids\, the uncertainty due to unresolved small-scale m
otions has to be taken into account as well as the loss of accuracy due to
poorly resolved spatial derivatives. In this presentation\, we discuss ho
w data assimilation methods can be used to derive data-driven stochastic f
orcing for coarse computational GFD. We will show that a straightforward a
lgorithm\, based on several simplifying assumptions\, already leads to qua
litatively accurate outcomes at strongly reduced computational costs.\n
LOCATION:https://researchseminars.org/talk/cam/3/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Akash Sharma (Chalmers University of Technology and University of
Gothenburg)
DTSTART:20230913T111500Z
DTEND:20230913T120000Z
DTSTAMP:20260416T215336Z
UID:cam/4
DESCRIPTION:Title: Rand
om walks for approximating boundary value problems\nby Akash Sharma (
Chalmers University of Technology and University of Gothenburg) as part of
CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nWe will present numer
ical method to simulate reflected stochastic differential equations. We ge
neralize this algorithm to approximately solve linear Robin boundary value
problems via their probabilistic representations. In next part of the tal
k\, we will present numerical schemes to simulate confined Langevin dynami
cs which results in approximate solution of specular boundary value proble
ms. We obtain rate of convergence of these algorithms and verify them with
numerical experiments. This is a joint work with Prof. Benedict Leimkuhle
r (University of Edinburgh) and Prof. Michael V. Tretyakov (University of
Nottingham).\n
LOCATION:https://researchseminars.org/talk/cam/4/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Ioanna Motschan-Armen (Chalmers University of Technology and Unive
rsity of Gothenburg)
DTSTART:20230927T111500Z
DTEND:20230927T120000Z
DTSTAMP:20260416T215336Z
UID:cam/5
DESCRIPTION:Title: Eule
r-Maruyama approximations of the stochastic heat equation on the sphere\nby Ioanna Motschan-Armen (Chalmers University of Technology and Univers
ity of Gothenburg) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAb
stract\nThe stochastic heat equation on the sphere driven by additive isot
ropic Wiener\nnoise is approximated by a spectral method in space and forw
ard and backward Euler–\nMaruyama schemes in time. The spectral approxim
ation is based on a truncation of the series\nexpansion with respect to th
e spherical harmonic functions. Optimal strong convergence rates\nfor a gi
ven regularity of the initial condition and driving noise are derived for
the Euler–\nMaruyama methods. Besides strong convergence\, convergence o
f the expectation and second\nmoment is shown\, where the approximation of
the second moment converges with twice the\nstrong rate. Numerical simula
tions confirm the theoretical results.\nThis is joint work with Annika Lan
g.\n
LOCATION:https://researchseminars.org/talk/cam/5/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Mohammad Asadzadeh (Chalmers & University of Gothenburg)
DTSTART:20231011T111500Z
DTEND:20231011T120000Z
DTSTAMP:20260416T215336Z
UID:cam/6
DESCRIPTION:Title: On N
itsche approach for a finite element scheme for Maxwell equations\nby
Mohammad Asadzadeh (Chalmers & University of Gothenburg) as part of CAM se
minar\n\nLecture held in MV:L14.\n\nAbstract\nWe show improved convergence
for a $h-p$\, streamline diffusion (SD)\, Nitsche's scheme for the Vlasov
-Maxwell (VM) system. The standard Galerkin for VM equations\, as 1st orde
r hyperbolic\, suffers from the draw-back of poor convergence. We have imp
roved this convergence rate using: \n\n(i) The SD method that adds artific
ial diffusion to the system.\n\n(ii) The $h-p$ approach to gain adaptivity
feature. \n\n(iii) Combined\, differentiated\, Maxwell equations to rende
r the first order hyperbolic system to a second order hyperbolic equation
(not applicable to Vlasov part). \n\n(iv) Add of {\\sl symmetry} and {\\sl
penalty} terms to reach final step of Nitsche's scheme.\n\nNumerical exam
ples are justifying the theory.\n
LOCATION:https://researchseminars.org/talk/cam/6/
END:VEVENT
BEGIN:VEVENT
SUMMARY:TBA
DTSTART:20231018T111500Z
DTEND:20231018T120000Z
DTSTAMP:20260416T215336Z
UID:cam/8
DESCRIPTION:by TBA as part of CAM seminar\n\nLecture held in MV:L14.\nAbst
ract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/8/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Institutionsdag
DTSTART:20231025T111500Z
DTEND:20231025T120000Z
DTSTAMP:20260416T215336Z
UID:cam/9
DESCRIPTION:by Institutionsdag as part of CAM seminar\n\nLecture held in M
V:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/9/
END:VEVENT
BEGIN:VEVENT
SUMMARY:TBA
DTSTART:20231101T121500Z
DTEND:20231101T130000Z
DTSTAMP:20260416T215336Z
UID:cam/10
DESCRIPTION:by TBA as part of CAM seminar\n\nLecture held in MV:L14.\nAbst
ract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/10/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Luigi Romano (Chalmers University of Technology)
DTSTART:20231108T121500Z
DTEND:20231108T130000Z
DTSTAMP:20260416T215336Z
UID:cam/11
DESCRIPTION:Title: Fin
ite element modelling of linear rolling contact problems\nby Luigi Rom
ano (Chalmers University of Technology) as part of CAM seminar\n\nLecture
held in MV:L14.\n\nAbstract\nThis Master's thesis deals with the numerical
approximation of linear hyperbolic problems appearing in rolling contact
mechanics. First\, the existence and uniqueness of strict solutions to the
considered equations\, which contain nonlocal and boundary terms\, are an
alysed within the framework provided by the semigroup theory. Then\, the s
pace semi-discrete problem is formulated using the discontinuous Galerkin
finite element method (DGMs)\, by replacing the unbounded operator appeari
ng in the abstract formulation with a finite-dimensional one. Quasi-optima
l error convergence is obtained for the space semi-discrete scheme by intr
oducing upwind regularisation. Time discretisation is then achieved by rel
ying on explicit first and second-order Runge-Kutta algorithms (RK1 and RK
2\, respectively)\, yielding quasi-optimal convergence in time owing to ce
rtain refined CFL conditions. In particular\, the considered RK2 schemes c
over the explicit midpoint method\, Heun's second-order method\, and Ralst
on's method.\n
LOCATION:https://researchseminars.org/talk/cam/11/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Moritz Hauck (Chalmers and GU)
DTSTART:20231115T121500Z
DTEND:20231115T130000Z
DTSTAMP:20260416T215336Z
UID:cam/12
DESCRIPTION:Title: Gua
ranteed lower energy bounds for the Gross-Pitaevskii problem using mixed f
inite elements\nby Moritz Hauck (Chalmers and GU) as part of CAM semin
ar\n\nLecture held in MV:L14.\n\nAbstract\nIn this talk\, we present a low
est order Raviart-Thomas finite element discretization that provides guara
nteed lower bounds on the ground state energy of the nonlinear Gross-Pitae
vskii problem. We emphasize that due to their conformity\, classical discr
etization methods such as the $\\mathcal P^1$ or $\\mathcal Q^1$ finite el
ement methods can only provide upper bounds on the ground state energy. Fu
rthermore\, we establish an a priori error analysis for the Raviart-Thomas
discretization of the Gross-Pitaevskii problem. Optimal convergence rates
are shown for the primary and dual variables as well as for the eigenvalu
e and energy approximations.\n
LOCATION:https://researchseminars.org/talk/cam/12/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Johannes Borgqvist (University of Oxford\, Chalmers and GU)
DTSTART:20231122T121500Z
DTEND:20231122T130000Z
DTSTAMP:20260416T215336Z
UID:cam/13
DESCRIPTION:Title: Lie
symmetries for constructing\, selecting and analysing mechanistic models
in mathematical biology\nby Johannes Borgqvist (University of Oxford\,
Chalmers and GU) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbs
tract\nGiven the abundance of experimental data\, two of the most fundamen
tal questions in mechanistic modelling of biological data concern model co
nstruction and model selection. A common type of data is time series data
describing how some quantity\, e.g. population size or protein abundance\,
changes over time. Given such a time series\, it is often possible to con
struct numerous candidate mechanistic models consisting of ordinary differ
ential equations based on physical principles encoding distinct biological
hypotheses. Worse still\, numerous candidate models of the same time seri
es often describe the same data equally well\, and thus they cannot be dis
tinguished based on their fit to data. In these situations it is therefore
difficult to select one candidate model and thereby infer a biological me
chanism underlying the biological data. In this presentation\, we tackle t
he two fundamental problems of model construction and model selection by m
eans of Lie symmetries (or simply just symmetries) of ordinary differentia
l equations. These are\, simply put\, (one parameter pointwise) transforma
tions known as $\\mathcal{C}^{\\infty}$ diffeomorphisms which map a soluti
on curve to another solution curve. Symmetries are commonly used in mathem
atical physics and they are the basis for numerous Nobel prizes but they a
re almost unheard of in mathematical biology.\n\nTo solve the classical mo
del selection problem\, we have developed and implemented a methodology fo
r model selection based on symmetries. We implement this framework on actu
al experimental data describing the age-related increase in cancer risk. I
mportantly\, we infer experimentally validated hypotheses underlying diffe
rent cancer types using the symmetry based framework which the standard me
thodology based on model fitting fails to do. \n\nThereafter\, we switch f
ocus to model construction in the context of travelling wave models of col
lective cell migration. These models consists of a single second order ODE
describing how the population density $u(z)$ changes with respect to a tr
avelling wave variable $z=x-ct$ where the constant $c$ is referred to as t
he wave speed. Moreover\, certain such models of reaction diffusion type a
s well as other models with density dependent diffusion are known to have
specific analytical solutions of a simple form for certain wave speeds. Th
ese analytical solutions have been obtained by means of ansätze based on
series expansions\, and using these methods it is difficult to define the
class of models which have simple analytical solutions. To tackle this pro
blem\, we consider a set of symmetries referred to as a Lie Algebra consis
ting of two symmetries that has been used to find analytical solutions of
a second order ODE encapsulating numerous oscillatory models such as the v
an der Pol oscillator. Based on differential invariants\, we derive the mo
st general class of models for which this Lie Algebra is manifest. Thereaf
ter\, we implement Lie's algorithm based on step-wise integration in order
to demonstrate how first integrals and (if possible) analytical solutions
of all ODEs in our class of models are obtained. Using this general class
of models\, we construct a sub-class of models characterised by the previ
ously mentioned simple analytical solution. Lastly\, we demonstrate how th
is sub-class encapsulates the previously known models with analytical solu
tions and we quantify the action of the symmetries in this Lie algebra on
these analytical solutions. In total\, this work demonstrates how classes
of mechanistic models can be constructed based on mathematical properties
encoded by a Lie Algebra in contrast to the standard way of model construc
tion based on physical assumptions that are hard to validate.\n
LOCATION:https://researchseminars.org/talk/cam/13/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Michael Roop (Chalmers and GU)
DTSTART:20231129T121500Z
DTEND:20231129T130000Z
DTSTAMP:20260416T215336Z
UID:cam/14
DESCRIPTION:Title: Lie
-Poisson methods for incompressible magnetohydrodynamics on the sphere
\nby Michael Roop (Chalmers and GU) as part of CAM seminar\n\nLecture held
in MV:L14.\n\nAbstract\nWe present a novel structure preserving numerical
method for Lie-Poisson systems on the dual of semidirect product Lie alge
bras. The method fully preserves the underlying geometry\, namely the Lie-
Poisson structure and all the Casimirs\, and nearly preserves the Hamilton
ian function. We illustrate the method on two models describing the motion
of magnetic fluids\, the equations of incompressible magnetohydrodynamics
\, and the Alfvén wave turbulence equations. For the latter case\, we rev
eal the formation of large scale quasi-periodic vortex blob dynamics.\n\nT
his is a joint work with Klas Modin.\n
LOCATION:https://researchseminars.org/talk/cam/14/
END:VEVENT
BEGIN:VEVENT
SUMMARY:TBA
DTSTART:20231206T121500Z
DTEND:20231206T130000Z
DTSTAMP:20260416T215336Z
UID:cam/15
DESCRIPTION:by TBA as part of CAM seminar\n\nLecture held in MV:L14.\nAbst
ract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/15/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Tommy Svensson and Anders Logg (Chalmers)
DTSTART:20231213T121500Z
DTEND:20231213T130000Z
DTSTAMP:20260416T215336Z
UID:cam/16
DESCRIPTION:Title: Sus
tainable Urban C-V2X with Intelligent Radio Environment Twinning\nby T
ommy Svensson and Anders Logg (Chalmers) as part of CAM seminar\n\nLecture
held in MV:L14.\n\nAbstract\nUbiquitous connectivity to power Information
and communication technologies (ICTs) is crucial for the modern society.
Besides\, ICT also has an important role for accelerating economic process
es that enable sustainability. As one part of ICT in smart and sustainable
cities\, cellular vehicle to everything (C-V2X)-based automated driving w
ill be crucial for fully autonomous or remotely operated vehicles in urban
environments. To this end\, different communication technologies have bee
n developed to improve the efficiency and reliability of the network perfo
rmance\, such as millimeter wave communications and reconfigurable intelli
gent surface (RIS). However\, challenges remain especially in urban areas
where the performance is normally hard to be evaluated before deployments.
Fortunately\, digital twinning can embed artificial intelligence and ICT
to provide a digital replica of real-life environments. With specific char
acteristics\, the replica is almost a cloned version of the original syste
m\, and is able to constantly update the properties with real-time data fr
om sensors. In this presentation\, we will present recent progress in our
AoA-ICT Seed Project on programmable\, and environmental-suitable communic
ation paradigm for C-V2X in dense city environments empowered by digital t
winning.\n
LOCATION:https://researchseminars.org/talk/cam/16/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Eric Lindström (Chalmers and GU)
DTSTART:20240226T121500Z
DTEND:20240226T130000Z
DTSTAMP:20260416T215336Z
UID:cam/17
DESCRIPTION:Title: Mul
tidimensional Coefficent Inverse Problems for Maxwellian Systems in Conduc
tive Media with Applications in Medical Imaging\nby Eric Lindström (C
halmers and GU) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstr
act\nThe talk will present a new method for medical imaging in the context
of coefficient inverse problems (CIPs) in maxwellian systems. The main ai
m is to find potential tumors in human breasts by reconstructing the value
and shape of spatial functions describing the dielectric properties of va
rious tissue types. The algorithms used for the hybrid method will be intr
oduced\, together with some theoretical results which discuss the well-pos
edness of the underlying problem\, as well as the convergence of the recon
structing method. We will also look at numerical results achieved by apply
ing our method to anatomically realistic data.\n
LOCATION:https://researchseminars.org/talk/cam/17/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Andreas Rupp (Lappeenranta-Lahti University of Technology)
DTSTART:20240219T121500Z
DTEND:20240219T130000Z
DTSTAMP:20260416T215336Z
UID:cam/18
DESCRIPTION:Title: Par
tial differential equations on hypergraphs and networks of surfaces\nb
y Andreas Rupp (Lappeenranta-Lahti University of Technology) as part of CA
M seminar\n\nLecture held in MV:L14.\n\nAbstract\nAlbeit many physical\, s
ociological\, engineering\, and economic processes have been described by
partial differential equations posed on domains which cannot be described
as subsets of linear spaces or smooth manifolds\, there is still a lack of
mathematical tools and general purpose software specifically addressing t
he challenges arising from the discretization of these models.\n\nThis pre
sentation establishes a general approach to formulate partial differential
equations (PDEs) on networks of (hyper)surfaces\, referred to as hypergra
phs. Such PDEs consist of differential expressions with respect to all hyp
eredges (surfaces) and compatibility conditions on the hypernodes (joints\
, intersections of surfaces). These compatibility conditions ensure conser
vation properties (in case of continuity equations) or incorporate other p
roperties – motivated by physical or mathematical modeling. We illuminat
e how to discretize such equations numerically using hybrid discontinuous
Galerkin (HDG) methods. These methods consist of local solvers (encoding t
he differential expressions on hyperedges) and a global compatibility cond
ition (related to our hypernode conditions). We complement the physically
motivated compatibility conditions by a derivation through a singular limi
t analysis of thinning structures yielding the same results.\n
LOCATION:https://researchseminars.org/talk/cam/18/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Brynjulf Owren (NTNU Trondheim)
DTSTART:20240212T121500Z
DTEND:20240212T130000Z
DTSTAMP:20260416T215336Z
UID:cam/19
DESCRIPTION:Title: Sta
bility of numerical methods on Riemannian manifolds\nby Brynjulf Owren
(NTNU Trondheim) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbs
tract\nStability of numerical integrators play a crucial role in approxima
ting the flow of differential equations. Issues related to convergence and
step size limitations have been successfully resolved by studying the sta
bility properties of numerical schemes. Stability also plays a role in the
existence and uniqueness to the solution of the nonlinear algebraic equat
ions that need to be solved in each time step for an implicit method.\nHow
ever\, very little has up to now been known about stability properties of
numerical methods on manifolds\, such as Lie group integrators. An interes
t in these questions has recently been sparked by the efforts in construct
ing ODE based neural networks that are robust against adversarial attacks.
In this talk we shall discuss a new framework for B-stability on Riemanni
an manifolds. A method is B-stable if the numerical method exhibits a non-
expansive behaviour in the Riemannian distance measure when applied to pro
blems which have non-expansive solutions.\nWe shall in particular see how
the sectional curvature of the manifold plays a role\, and show some surpr
ising results regarding the non-uniqueness of geodesic implicit integrator
s for positively curved spaces.\n
LOCATION:https://researchseminars.org/talk/cam/19/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Andrea Papini (Chalmers and GU)
DTSTART:20240304T121500Z
DTEND:20240304T130000Z
DTSTAMP:20260416T215336Z
UID:cam/20
DESCRIPTION:Title: Tur
bulence enhancement of coagulating processes\nby Andrea Papini (Chalme
rs and GU) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\n
We present and investigate the collision-coalescence process of particles
in the presence of a fluid velocity field\, examining the relationship bet
ween flow properties and enhanced coagulation. Our research focuses on two
main aspects. Firstly\, we propose a novel modeling approach for turbulen
t fluid at small scales\, employing a Gaussian random field with non-trivi
al spatial covariance. Secondly\, we derive rigorous partial differential
equations (PDEs) and stochastic partial differential equations (SPDEs) fro
m this model\, capturing the physical characteristics of particles suspend
ed in the fluid. From an Eulerian perspective\, we analyze a kinetic parti
cle system subjected to environmental transport noise. Specifically\, we r
igorously study a modified version of Smoluchowski’s coagulation equatio
n\, which incorporates velocity dependence akin to the Boltzmann equation.
By utilizing techniques rooted in unbounded elliptic semigroup theory and
weighted Sobolev space inequalities\, we establish the existence and uniq
ueness of classical solutions for the case of a spatially homogeneous init
ial distribution. Moreover\, from a Lagrangian viewpoint\, we employ this
particle system to gain insights into the collision rate at a steady state
for particles uniformly distributed within a medium. Considering a partic
le-fluid model\, we perform two scaling limits. The first limit\, involvin
g the number of particles\, yields a stochastic Smoluchowski-type system\,
with the turbulent velocity field still governed by a noise stochastic pr
ocess. The second scaling limit pertains to the parameters of the noise\,
specifically targeting the direction associated with small-scale turbulenc
e. This limit leads to a deterministic equation with eddy dissipation in t
he velocity variable. We conduct numerical simulations of this equation sy
stem and demonstrate the influence of turbulence on rain formation. Our qu
alitative findings reveal a steady increase in coagulation efficiency with
escalating turbulent kinetic energy of the fluid. Additionally\, we obser
ve a power-law decay over time and in relation to the turbulence parameter
. Furthermore\, we recover fundamental laws governing the collision rate a
nd relative velocity of moving particles in the high Stokes number regime.
\n
LOCATION:https://researchseminars.org/talk/cam/20/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Verena Schwarz (University of Klagenfurt)
DTSTART:20240318T121500Z
DTEND:20240318T130000Z
DTSTAMP:20260416T215336Z
UID:cam/21
DESCRIPTION:Title: Hig
her-order approximation and optimality for jump-diffusion SDEs with discon
tinuous drift\nby Verena Schwarz (University of Klagenfurt) as part of
CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nIn this talk we consi
der the approximation of jump-diffusion stochastic differential equations
with discontinuous drift\, possibly degenerate diffusion coefficient\, and
Lipschitz continuous jump coefficient. We present a jump-adapted higher-o
rder scheme\, the so-called transformation-based jump-adapted quasi-Milste
in scheme. For this scheme\, we provide a complete error analysis: We prov
e convergence order $3/4$ in $L^p$ for $p\\in[1\,\\infty)$. Further\, we p
rovide lower error bounds for non-adaptive and jump-adapted approximation
schemes of order $3/4$ in $L^1$. This yields optimality of the transformat
ion-based jump-adapted quasi-Milstein scheme.\n\nThis is joint work with P
awel Przybylowicz and Michaela Szölgyenyi.\n
LOCATION:https://researchseminars.org/talk/cam/21/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Stefano Di Giovacchino (University of L'Aquila)
DTSTART:20240408T111500Z
DTEND:20240408T120000Z
DTSTAMP:20260416T215336Z
UID:cam/22
DESCRIPTION:Title: Sto
chastic backward error analysis: application to Hamiltonian systems and op
timization algorithms\nby Stefano Di Giovacchino (University of L'Aqui
la) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nBackwar
d error analysis is a powerful tool in order to capture the long-term beha
viour of numerical integrators. In this talk\, we address our attention on
providing a backward error analysis (both in the strong and weak sense) f
or classes of numerical methods. From one hand\, the attention will be dev
oted to symplectic methods and Poisson integrators for stochastic Hamilton
ian and Poisson systems. Here\, we present strategies for deriving stochas
tic modified equations for the aforementioned integrators and we study the
m for obtaining long-term estimates on the Hamiltonian errors along the nu
merical dynamics.\nFrom the other hand\, the weak backward error approach
will be developed towards stochastic optimization algorithms\, with the ai
m of gaining insights of their behaviour.\nThis talk is based on joint res
earches with Raffaele D'Ambrosio (University of L'Aquila)\, Desmond J. Hig
ham and Konstantios C. Zygalakis (University of Edinburgh).\n
LOCATION:https://researchseminars.org/talk/cam/22/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Fanny Seizilles (University of Cambride)
DTSTART:20240415T111500Z
DTEND:20240415T120000Z
DTSTAMP:20260416T215336Z
UID:cam/23
DESCRIPTION:Title: The
Bayesian approach to inverse Robin problems\nby Fanny Seizilles (Univ
ersity of Cambride) as part of CAM seminar\n\nLecture held in MV:L14.\n\nA
bstract\nWe investigate the Bayesian approach to certain elliptic boundary
value problems of determining a Robin coefficient on a hidden part of the
boundary from Cauchy data on the observable part. Such a nonlinear invers
e problem arises naturally in the initialisation of large-scale ice sheet
models. In this talk we will specifically focus on the computational routi
ne to estimate posterior densities for the Robin coefficient.\n\n\nThe Bay
esian approach is motivated for a prototypical Robin inverse problem by sh
owing that the posterior mean converges in probability to the data-generat
ing ground truth as the number of observations increases. Related to the s
tability theory for inverse Robin problems\, a logarithmic convergence rat
e for Sobolev-regular Robin coefficients is established\, whereas for anal
ytic coefficients an algebraic rate can be attained. Our numerical results
on synthetic data illustrate the convergence property in two observation
settings. (Joint work with Aksel Kaastrup Rasmussen\, Ieva Kazlauskaite an
d Mark Girolami).\n
LOCATION:https://researchseminars.org/talk/cam/23/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Zheng Zhao (Uppsala University)
DTSTART:20240205T121500Z
DTEND:20240205T130000Z
DTSTAMP:20260416T215336Z
UID:cam/24
DESCRIPTION:Title: Mom
ent quadrature for stochastic filtering\nby Zheng Zhao (Uppsala Univer
sity) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nStoch
astic filtering is an important estimation problem for time series. In thi
s talk\, we present a class of filters that represent the filtering distri
butions by their moments. The key enablement is a quadrature method that u
ses orthonormal polynomials spanned by the moments. We show that this mome
nt-based filter is asymptotically exact in the order of moments\, and that
the filter is also computationally efficient and is in line with the stat
e of the art.\n
LOCATION:https://researchseminars.org/talk/cam/24/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Axel Ringh (Chalmers and GU)
DTSTART:20240506T111500Z
DTEND:20240506T120000Z
DTSTAMP:20260416T215336Z
UID:cam/25
DESCRIPTION:Title: Gai
n and phase type multipliers for structured feedback robustness\nby Ax
el Ringh (Chalmers and GU) as part of CAM seminar\n\nLecture held in MV:L1
4.\n\nAbstract\nOne of the most fundamental problems in control theory is
feedback stability analysis. That is\, the problem of determining if two s
ystems interconnected via feedback will be stable. In the case of linear t
ime-invariant systems\, under mild conditions the solvability of a set of
linear matrix inequalities (LMIs) is a both necessary and sufficient condi
tion for stability. Nevertheless\, models of reality are always imperfect\
, and in robust stability analysis one therefore instead consider the prob
lem if a feedback interconnection between a nominal system and a set of un
certainties is stable for all uncertainties in the set. In this talk\, I w
ill present new results on that robustness to certain forms of structured
uncertainties is equivalent with the existence of certain forms of structu
red solutions to the LMIs. The talk is aimed to be self-contained\; no pri
or knowledge on control theory is needed\, and all relevant concepts will
be introduced and explained.\n
LOCATION:https://researchseminars.org/talk/cam/25/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Klas Modin (Chalmers and GU)
DTSTART:20240129T121500Z
DTEND:20240129T130000Z
DTSTAMP:20260416T215336Z
UID:cam/26
DESCRIPTION:Title: The
reversibility paradox in matrix hydrodynamics\nby Klas Modin (Chalmer
s and GU) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nS
ome time ago\, Milo Viviani and myself unveiled numerical simulations of i
ncompressible 2-D hydrodynamics on the sphere indicating a connection betw
een the long-time behavior of 2-D Euler equations and integrability condit
ions for "blob dynamics". After presenting these results\, I was asked an
insightful question:\n\nThe phase space underlying the model in the simula
tions is compact. Because the dynamics in the model is also Hamiltonian\,
we have Poincaré recurrence. But the dynamics in the simulations\, leadin
g to blob formations\, seem contractive. Isn't the mechanism for blob form
ations instead induced by fictitious dissipation\, introduced via the nume
rical time-discretization?\n\nI didn’t have a good answer at the time\,
but the question stayed with me. Today I have an answer\, which is the sub
ject of this talk.\n
LOCATION:https://researchseminars.org/talk/cam/26/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Eddie Wadbro (Karlstad University)
DTSTART:20240422T111500Z
DTEND:20240422T120000Z
DTSTAMP:20260416T215336Z
UID:cam/27
DESCRIPTION:Title: Mat
erial distribution topology optimization for boundary-effect-dominated pro
blems\nby Eddie Wadbro (Karlstad University) as part of CAM seminar\n\
nLecture held in MV:L14.\n\nAbstract\nIn the classical design optimization
using the material distribution method (density-based topology optimizati
on)\, a material indicator function represents the presence or absence of
material within the domain. The first part of this talk provides an introd
uction to material distribution topology optimization with an emphasis on
mathematical morphology\, non-linear filters\, and length scale control.\n
\nTo use the material distribution approach for boundary-effect-dominated
problems\, we need to identify the boundary of the design at each iteratio
n\; this talk discusses two methods to achieve this. The first is to use a
boundary strip indicator function defined on the elements of the computat
ional mesh. The second is to use a boundary indicator function defined on
the mesh faces (edges in 2D and facets in 3D). The second part of my prese
ntation covers the main ideas behind both approaches and showcases results
from two applications\, one suitable for each approach.\n
LOCATION:https://researchseminars.org/talk/cam/27/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Ragnar Winther
DTSTART:20240429T111500Z
DTEND:20240429T120000Z
DTSTAMP:20260416T215336Z
UID:cam/28
DESCRIPTION:Title: Wha
t about $p$?\nby Ragnar Winther as part of CAM seminar\n\nLecture held
in MV:L14.\n\nAbstract\nThe title of this talk refers to a question frequ
ently asked by Ivo Babuska\, from the late 1980s and on\, following variou
s talks \non finite element methods. More precisely\, what can we say abou
t the properties of finite element methods as we raise the polynomial degr
ee? Even today the so called $p$-method is less understood\,\nand the corr
esponding analysis is less canonical\, than the traditional approach of me
sh refinement\, i.e.\, the $h$-method.\nIn recent years Rick Falk and I ha
ve developed a theory which represents a new tool to analyze finite elemen
t methods of high polynomial degree\,\nwhich we refer to as the bubble t
ransform. The key idea is to construct a decomposition into local bubbles
which \nsimultaneously covers all possible polynomial degrees.\nThe purpos
e of this talk is to give a review of this theory\, and to discuss \npoten
tial applications.\n
LOCATION:https://researchseminars.org/talk/cam/28/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Easter Monday
DTSTART:20240401T111500Z
DTEND:20240401T120000Z
DTSTAMP:20260416T215336Z
UID:cam/29
DESCRIPTION:by Easter Monday as part of CAM seminar\n\nLecture held in MV:
L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/29/
END:VEVENT
BEGIN:VEVENT
SUMMARY:TBC
DTSTART:20240527T111500Z
DTEND:20240527T120000Z
DTSTAMP:20260416T215336Z
UID:cam/30
DESCRIPTION:by TBC as part of CAM seminar\n\nLecture held in MV:L14.\nAbst
ract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/30/
END:VEVENT
BEGIN:VEVENT
SUMMARY:TBC
DTSTART:20240610T111500Z
DTEND:20240610T120000Z
DTSTAMP:20260416T215336Z
UID:cam/31
DESCRIPTION:by TBC as part of CAM seminar\n\nLecture held in MV:L14.\nAbst
ract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/31/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Christian Lubich (University of Tübingen)
DTSTART:20240409T111500Z
DTEND:20240409T120000Z
DTSTAMP:20260416T215336Z
UID:cam/32
DESCRIPTION:Title: Reg
ularized dynamical nonlinear parametric approximation\nby Christian Lu
bich (University of Tübingen) as part of CAM seminar\n\nLecture held in M
V:L14.\n\nAbstract\nThis talk is about the numerical approximation of solu
tions to initial value problems of high-dimensional ordinary differential
equations or evolutionary partial differential equations such as the Schr\
\"odinger equation by nonlinear parametrizations $u(t)=\\Phi(q(t))$ with t
ime-dependent parameters $q(t)$\, which are to be determined in the comput
ation. Our motivation comes from approximations by multiple Gaussians in q
uantum dynamics\, by tensor networks\, and by neural networks. In all thes
e cases\, the parametrization is typically irregular: the derivative $\\Ph
i'(q)$ can have arbitrarily small singular values and may have varying ran
k. The talk is about approximation results for a regularized approach\, wh
ich can still be successfully applied in such irregular situations\, even
if it runs counter to the basic principle in numerical analysis to avoid s
olving ill-posed subproblems when aiming for a stable algorithm.\nThe talk
is based on joint work with Jörg Nick\, Caroline Lasser and Michael Feis
chl.\n
LOCATION:https://researchseminars.org/talk/cam/32/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Malin Rau (Universität Hamburg)
DTSTART:20240311T121500Z
DTEND:20240311T130000Z
DTSTAMP:20260416T215336Z
UID:cam/33
DESCRIPTION:Title: Asy
nchronous Opinion Dynamics in Social Networks\nby Malin Rau (Universit
ät Hamburg) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract
\nOpinion spreading in society decides the fate of elections\, the success
of products\, and the impact of political or social movements.\nA promine
nt model to study opinion formation processes is due to Hegselmann and Kra
use. It has the distinguishing feature that stable states do not necessari
ly show consensus\, i.e.\, the population of agents might not agree on the
same opinion.\n\nWe focus on the social variant of the Hegselmann-Krause
model. There are $n$ agents\, which are connected by a social network. The
ir opinions evolve in an iterative\, asynchronous process in which agents
are activated one after another at random. When activated\, an agent adopt
s the average of the opinions of its neighbors having a similar opinion (w
here similarity of opinions is defined using a parameter $\\varepsilon$).
Thus\, the set of influencing neighbors of an agent may change over time.
To the best of our knowledge\, social Hegselmann-Krause systems with async
hronous opinion updates have only been studied with the complete graph as
social network.\n\nWe show that such opinion dynamics are guaranteed to co
nverge for any social network. We provide an upper bound of $\\mathcal{O}(
n|E|^2 (\\varepsilon/\\delta)^2)$ on the expected number of opinion update
s until convergence to a stable state\, where $|E|$ is the number of edges
of the social network\, and $\\delta$ is a parameter of the stability con
cept. For the complete social network\, we show a bound of $\\mathcal{O}(n
^3(n^2 + (\\varepsilon/\\delta)^2))$ that represents a major improvement o
ver the previously best upper bound of $\\mathcal{O}(n^9 (\\varepsilon/\\d
elta)^2)$.\n
LOCATION:https://researchseminars.org/talk/cam/33/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Ioanna Motschan-Armen (Chalmers & GU)
DTSTART:20240617T091500Z
DTEND:20240617T100000Z
DTSTAMP:20260416T215336Z
UID:cam/34
DESCRIPTION:Title: App
roximation of semilinear stochastic heat equations on the sphere\nby I
oanna Motschan-Armen (Chalmers & GU) as part of CAM seminar\n\nLecture hel
d in MV:H12.\n\nAbstract\nStochastic partial differential equations are us
ed to describe various physical processes that are perturbed by noise. Som
e of those occur on curved surfaces\, for example spheres. In this talk se
milinear stochastic heat equations with additive noise on the unit sphere
are considered. Approximations in space and time are presented in order to
simulate and analyse solutions. The space approximation is derived using
the spectral method\, with spherical harmonic functions. In order to obtai
n time discretization on an equidistant time grid the Euler--Maruyama sche
me is applied. For the semilinear stochastic heat equations on the sphere
with additive isotropic Wiener noise\, strong convergence rates in space a
nd time are derived\, taking regularity of the initial condition and the d
riving noise into account. Furthermore convergence of the expectation and
the second moment is analysed for the corresponding linear equation. The t
heoretical results are confirmed by numerical simulations.\n\nMidterm-Semi
nar\n
LOCATION:https://researchseminars.org/talk/cam/34/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Charles-Edouard Bréhier (Université de Pau et des Pays de l'Adou
r)
DTSTART:20241007T111500Z
DTEND:20241007T120000Z
DTSTAMP:20260416T215336Z
UID:cam/35
DESCRIPTION:Title: Asy
mptotic error analysis of stochastic optimization schemes\nby Charles-
Edouard Bréhier (Université de Pau et des Pays de l'Adour) as part of CA
M seminar\n\nLecture held in MV:L14.\n\nAbstract\nStochastic optimization
algorithms are nowadays widely used\, especially in the machine learning c
ommunity. In this talk\, we study a class of stochastic optimization schem
es which are perturbations of gradient descent algorithms.\nWe perform a r
igorous analysis of the convergence\, with proofs of error bounds with res
pect to the time-step size\, in the large time regime\, in the case of str
ongly convex objective functions. The error bounds follow from an interpre
tation of the schemes in terms of deterministic and stochastic modified eq
uations\, and using tools from weak error analysis of numerical methods fo
r stochastic differential equations.\n
LOCATION:https://researchseminars.org/talk/cam/35/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Mykola Tkachuk (Karlstad University)
DTSTART:20240923T111500Z
DTEND:20240923T120000Z
DTSTAMP:20260416T215336Z
UID:cam/36
DESCRIPTION:Title: Var
iational approaches to the analysis of structural response of fiber networ
k microstructures\nby Mykola Tkachuk (Karlstad University) as part of
CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nTwo approaches to stat
ic equilibrium of fiber network microstructures will be presented.\nOrient
ation-based averaging of network microstructures with non-affine kinematic
s associates the ensemble of fibers at a given material point with a unit
sphere of initial orientations (called also microsphere or representative
orientation space). A special micro-macro relation between these microscop
ic deformations and the macroscopic deformation gradient is derived by the
concept of maximal advance paths in the network. The averaged microscopic
free energy of the fibers is minimized subjected to this kinematic constr
aint. The derived model can be applied to networks of flexible polymer cha
ins (as found in elastomers)\, semiflexible or stiff filaments (elastin\,
collagen and other biopolymers and soft tissues). The model has been recen
tly extended to composite networks consisting of two or more distinguished
components. \nSliding of inextensible fibers in discrete network microstr
ucture of nonwoven materials is defined as a nonsmooth convex optimization
problem. The rate-independent sliding and the quasi-static equilibrium lo
ading of such network structure are constituted within the theory of stand
ard dissipative systems. A minimum principle for incremental potential is
formulated with respect to the displacement-based variables as a second-or
der cone programming (SOCP) problem. A pure complementary energy principle
is derived as the dual formulation in terms of stress-like variables. Bot
h SOCP problems can be solved numerically by interior-point methods.\n
LOCATION:https://researchseminars.org/talk/cam/36/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Workshop STINT: Sweden-China
DTSTART:20241014T111500Z
DTEND:20241014T120000Z
DTSTAMP:20260416T215336Z
UID:cam/37
DESCRIPTION:by Workshop STINT: Sweden-China as part of CAM seminar\n\nLect
ure held in MV:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/37/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Andrii Dmytryshyn (Örebro University)
DTSTART:20240902T111500Z
DTEND:20240902T120000Z
DTSTAMP:20260416T215336Z
UID:cam/38
DESCRIPTION:Title: Eig
enstructures of low-rank matrix polynomials\nby Andrii Dmytryshyn (Ör
ebro University) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbst
ract\nChallenging and intriguing mathematical problems involving matrix po
lynomials arise in various applications. These problems often revolve arou
nd the eigenstructures of the polynomials\, emphasizing the importance of
the eigenstructures. In this talk we consider the set of matrix polynomial
s of bounded rank and degree and describe the eigenstructures that these p
olynomials typically have\, so called generic eigenstructures. We also fin
d such generic eigenstructures for the sets of symmetric and skew-symmetri
c matrix polynomials. Notably\, these symmetries have drastic effect on ge
neric eigenstructures\, for example\, on whether we can anticipate the app
earance of eigenvalues in the eigenstructures or not. \nThis talk is prima
rily based on joint work with Froilán Dopico and Fernando De Téran.\n
LOCATION:https://researchseminars.org/talk/cam/38/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Alex (Oleksandr) Misiats (Virginia Commonwealth University)
DTSTART:20240819T111500Z
DTEND:20240819T120000Z
DTSTAMP:20260416T215336Z
UID:cam/39
DESCRIPTION:Title: Lin
ear Elastic Models of Shape Memory Alloys: Variational Perspective\nby
Alex (Oleksandr) Misiats (Virginia Commonwealth University) as part of CA
M seminar\n\nLecture held in MV:L14.\n\nAbstract\nIn this talk\, I will de
scribe a class of minimization problems arising in modelling shape memory
alloys. I will start with a shape memory material illustration\, followed
by a simple one dimensional model of it. Its extensions in 2D and 3D will
help us understand the energetic mechanism behind the formation of twin p
atterns in physical experiments. By means of sharp upper and lower bounds\
, we show that certain experimentally observed structures provide optimal
energy scaling law. Despite the fact that the problem is highly nonconvex
due to the presence of phase constraints and the singular perturbation\, i
n my talk\, I will describe a relaxation method which allows the use of th
e convex duality technique for the purpose of obtaining a sharp lower boun
d. I will also describe the extension of the linear elastic models of two
twins to model multiscale patterns\, which appear in recent physical exper
iments and involve mixing of four twins.\n
LOCATION:https://researchseminars.org/talk/cam/39/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Gabriel Velho (CentraleSupélec)
DTSTART:20240826T111500Z
DTEND:20240826T120000Z
DTSTAMP:20260416T215336Z
UID:cam/40
DESCRIPTION:Title: Sta
bilization and Optimal Control of Interconnected SDE - Scalar PDE System\nby Gabriel Velho (CentraleSupélec) as part of CAM seminar\n\nLecture
held in MV:L14.\n\nAbstract\nIn this talk\, I will present my work on the
optimal control and stabilization of a system composed of an interconnecte
d Stochastic Differential Equation (SDE) and a Scalar Partial Differential
Equation (PDE)\, with control applied at the boundary of the PDE. Such sy
stems can be found in practical scenarios\, such as in temperature control
of a building where heat must be transmitted from a distance.\nI will beg
in with an introduction to boundary control of PDEs\, emphasizing the back
stepping method—a powerful technique for designing boundary controllers
that stabilize PDEs. In particular\, we will see how backstepping can be e
mployed to stabilize a system of coupled hetero-directional transport equa
tions.\nNext\, we will examine the complete system where this PDE is conne
cted with an SDE at the boundary. I will show how backstepping transforms
this problem into an equivalent control problem for an input-delayed SDE\,
making it more tractable.\nFinally\, I will discuss the implications of i
nput delays in controlling SDEs with additive noise\, particularly how the
se delays impact system performance and the strategies we can use to mitig
ate these effects.\n
LOCATION:https://researchseminars.org/talk/cam/40/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Malin Rau (Chalmers & GU)
DTSTART:20240916T111500Z
DTEND:20240916T120000Z
DTSTAMP:20260416T215336Z
UID:cam/41
DESCRIPTION:Title: Sch
eduling with Many Shared Resources\nby Malin Rau (Chalmers & GU) as pa
rt of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nConsider the man
y shared resources scheduling problem where jobs have to be scheduled on i
dentical parallel machines with the goal of minimizing the makespan.\nHowe
ver\, each job needs exactly one additional shared resource in order to be
executed and hence prevents the execution of jobs that need the same reso
urce while being processed.\nPreviously\, an approximation ratio of asympt
otically 2 was the best known result for this problem.\nFurthermore\, a 6/
5-approximation for the case with only two machines was known as well as a
PTAS for the case with a constant number of machines.\nWe present a simpl
e and fast 5/3-approximation and a much more involved but still reasonable
1.5-approximation.\nFurthermore\, we provide a PTAS for the case with onl
y a constant number of machines\, which is arguably simpler and faster tha
n the previously known one\, as well as a PTAS with resource augmentation
for the general case.\nThe approximation schemes make use of the N-fold in
teger programming machinery\, which has found more and more applications i
n the field of scheduling recently.\nIt is plausible that the latter resul
ts can be improved and extended to more general cases.\nLastly\, we give a
n inapproximability result for the natural problem extension where each jo
b may need up to a constant number of different resources\, namely 3\, rul
ing out better than 5/4 approximations for that case.\n
LOCATION:https://researchseminars.org/talk/cam/41/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Arnout Franken (University of Twente)
DTSTART:20240930T111500Z
DTEND:20240930T120000Z
DTSTAMP:20260416T215336Z
UID:cam/42
DESCRIPTION:Title: Cas
imir-preserving numerical methods for global geophysical fluid dynamics\nby Arnout Franken (University of Twente) as part of CAM seminar\n\nLect
ure held in MV:L14.\n\nAbstract\nIn this talk\, I will present the develop
ment of novel structure-preserving numerical methods for simulating global
geophysical fluid dynamics. The methods are derived using a geometric app
roach to the quasi-geostrophic (QG) equations on a sphere\, revealing thei
r Lie-Poisson structure. This leads to energy and enstrophy-conserving num
erical methods that also preserve higher-order invariants\, known as Casim
irs. These methods enable long-time simulations and are applied to study g
eostrophic turbulence and its properties\, zonal jet formation\, and criti
cal latitudes in both single- and multi-layer models. I'll discuss both th
e development of the numerical methods and findings from simulations\, inc
luding insights into the role of geostrophic balance and stratification in
large-scale global fluid dynamics.\n
LOCATION:https://researchseminars.org/talk/cam/42/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Kristian Debrabant (University of Southern Denmark)
DTSTART:20241209T121500Z
DTEND:20241209T130000Z
DTSTAMP:20260416T215336Z
UID:cam/43
DESCRIPTION:Title: Opt
imal stochastic Runge-Kutta methods for the weak integration of stochastic
dynamics\nby Kristian Debrabant (University of Southern Denmark) as p
art of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nThe difficulty
of the creation of weak high order integrators for stochastic dynamics lie
s in the tedious calculations of order conditions. The original approaches
focused on adapting strong approximations\, mainly replacing the iterated
stochastic integrals by random variables that have the same moments. The
methods obtained this way are sub-optimal in their number of function eval
uations.\n\nIn this talk\, using a specific set of random Runge-Kutta coef
ficients\, we greatly reduce the number of order conditions for weak secon
d order integration of stochastic dynamics. The approach is successfully a
pplied to the creation of a collection of new simple stochastic Runge-Kutt
a methods of weak order two with an optimal number of stages.\n\nThe conte
nts of this talk are joint work with Anne Kværnø and Adrien Laurent.\n
LOCATION:https://researchseminars.org/talk/cam/43/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Johan Karlsson (KTH)
DTSTART:20241125T121500Z
DTEND:20241125T130000Z
DTSTAMP:20260416T215336Z
UID:cam/44
DESCRIPTION:Title: Str
uctured multi-marginal optimal transport: Applications\, theory\, and comp
utations\nby Johan Karlsson (KTH) as part of CAM seminar\n\nLecture he
ld in MV:L14.\n\nAbstract\nThe optimal mass transport problem is a classic
al problem in mathematics\, and dates back to 1781 and work by G. Monge wh
ere he formulated an optimization problem for minimizing the cost of trans
porting soil for construction of forts and roads. Historically the optimal
mass transport problem has been widely used in economics in\, e.g.\, plan
ning and logistics\, and was at the heart of the 1975 Nobel Memorial Prize
in Economic Sciences. In the last two decades there has been a rapid deve
lopment of theory and methods for optimal mass transport and the ideas hav
e attracted considerable attention in several economic and engineering fie
lds. These developments have led to a mature framework for optimal mass tr
ansport with computationally efficient algorithms that can be used to addr
ess many problems in applied mathematics.\n\n \n\nIn this talk\, I will gi
ve an overview of the multi-marginal optimal mass transport framework and
show how it can be applied to address and solve a range of problems in con
trol and estimation of multi-agent systems. This the optimal transport fra
mework allows for replacing the standard state space formalist\, where a s
tate evolve over time\, to a setting where instead densities or multi-agen
t systems evolve over time. In this setting we can formulate and solve a l
arge set of problems\, e.g.\, with given dynamics of the underlying agents
and multiple classes of agents\, nonlocal interactions\, or include const
raints between different time points such as origin destination constraint
s. We will also consider computational methods\, and motivated by Sinkhorn
's method for the standard optimal transport problems\, it can be shown th
at dual coordinate ascent is a computationally efficient approach for this
class of problems. \n\nIn this talk\, I will give an overview of the mult
i-marginal optimal mass transport framework and show how it can be applied
to address and solve a range of problems in control and estimation of mul
ti-agent systems. This the optimal transport framework allows for replacin
g the standard state space formalist\, where a state evolve over time\, to
a setting where instead densities or multi-agent systems evolve over time
. In this setting we can formulate and solve a large set of problems\, e.g
.\, with given dynamics of the underlying agents and multiple classes of a
gents\, nonlocal interactions\, or include constraints between different t
ime points such as origin destination constraints. We will also consider c
omputational methods\, and motivated by Sinkhorn's method for the standard
optimal transport problems\, it can be shown that dual coordinate ascent
is a computationally efficient approach for this class of problems.\n
LOCATION:https://researchseminars.org/talk/cam/44/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Gunilla Kreiss (Uppsala University)
DTSTART:20241104T121500Z
DTEND:20241104T130000Z
DTSTAMP:20260416T215336Z
UID:cam/45
DESCRIPTION:Title: Cut
FEM meets Finite Differences\nby Gunilla Kreiss (Uppsala University)
as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nThere is a
cut-FEM methodology with ghost penalty stabilization\, which can be applie
d to hyperbolic conservation laws. This technique is useful when consideri
ng problems with complex geometric features. Explicit time-stepping is pre
ferable for hyperbolic problems\, but even the standard DG and CG metods s
uffer from increasingly severe time-step restrictions as the order of the
method increases. For high order finite difference methods the time-step r
estriction is not at all as severe. In this talk we will explore possibili
ties of applying the cut-FEM methodology to finite difference methods. The
goal is to formulate a finite difference method that can be seen as a Gal
erkin method. Applying the cut-FEM methodology will then yield an immersed
boundary finite difference metod.\n
LOCATION:https://researchseminars.org/talk/cam/45/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Björn Müller (Chalmers & GU)
DTSTART:20241118T121500Z
DTEND:20241118T130000Z
DTSTAMP:20260416T215336Z
UID:cam/46
DESCRIPTION:Title: Iso
tropic Q-fractional Brownian motion on the sphere: regularity and fast sim
ulation\nby Björn Müller (Chalmers & GU) as part of CAM seminar\n\nL
ecture held in MV:L14.\n\nAbstract\nAs an extension of isotropic Gaussian
random fields and Q-Wiener processes on d-dimensional spheres\, isotropic
Q-fractional Brownian motion is introduced and sample Hölder regularity i
n space-time is shown depending on the regularity of the spatial covarianc
e operator Q and the Hurst parameter H. The processes are approximated by
a spectral method in space for which strong and almost sure convergence ar
e shown. The underlying sample paths of fractional Brownian motion are sim
ulated by circulant embedding or conditionalized random midpoint displacem
ent. Temporal convergence and computational complexity are numerically tes
ted\, the latter matching the complexity of simulating a Q-Wiener process
if allowing for a temporal error.\n
LOCATION:https://researchseminars.org/talk/cam/46/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Mika Persson (Chalmers & GU)
DTSTART:20241202T121500Z
DTEND:20241202T130000Z
DTSTAMP:20260416T215336Z
UID:cam/47
DESCRIPTION:Title: Dec
entralized control of teams of drones: A mixed collaborative/competitive g
ame\nby Mika Persson (Chalmers & GU) as part of CAM seminar\n\nLecture
held in MV:L14.\n\nAbstract\nSmall and inexpensive drones are increasingl
y used in surveillance\, reconnaissance\, and attack operations. Future dr
ones are expected to operate autonomously in large swarms\, posing challen
ges for traditional Surface-Based Air Defense (SBAD) systems\, which rely
on radar sensors and effectors like missiles or jamming. Swarms can overwh
elm SBAD systems due to their sheer numbers and economic asymmetry\, as dr
ones are significantly cheaper than defense systems. A potential counterme
asure is deploying defensive swarms of small drones with sensors and effec
tors\, necessitating optimized swarm behavior and capability evaluation. T
he proposed project focuses on decentralized control of such swarms using
game theory\, addressing challenges in operational goal representation\, s
ituational awareness through sensing and communication\, and managing unce
rtainty about enemy capabilities. These challenges are modeled by Partiall
y Observable Stochastic Games. Approximation techniques like Multi-Agent R
einforcement Learning are explored\, leveraging algorithms like MADDPG and
MAAC for mixed competitive and collaborative swarm-versus-swarm scenarios
. While related works address components of the problem\, they fall short
of addressing the full complexity\, particularly in handling unknown drone
numbers\, controlled sensing\, communication\, and intelligent adversarie
s in a competitive game-theoretic framework.\n
LOCATION:https://researchseminars.org/talk/cam/47/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Lucia Swoboda (Chalmers & GU)
DTSTART:20241111T121500Z
DTEND:20241111T130000Z
DTSTAMP:20260416T215336Z
UID:cam/48
DESCRIPTION:Title: On
a Galerkin-Bubnov variational formulation for the heat equation in anisotr
opic Sobolev spaces\, using the modified Hilbert transform\nby Lucia S
woboda (Chalmers & GU) as part of CAM seminar\n\nLecture held in MV:L14.\n
\nAbstract\nThe modified Hilbert transform arises naturally when consideri
ng a Galerkin-Bubnov variational formulation for the heat equation in anis
otropic Sobolev spaces. In this talk\, I will introduce the modified Hilbe
rt transform and its main properties. I will explain how this operator lea
ds to unique solvability of a variational formulation in anisotropic Sobol
ev spaces and to stability in the case of space-time tensor product discre
tization. Although the discrete inf-sup constant depending on the finite e
lement mesh parameter initially suggests a reduced order of convergence du
e to the structure of Céa's lemma\, optimal convergence is observed for a
large class of functions.\n
LOCATION:https://researchseminars.org/talk/cam/48/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Vincent Molin (Chalmers & GU)
DTSTART:20241028T121500Z
DTEND:20241028T130000Z
DTSTAMP:20260416T215336Z
UID:cam/49
DESCRIPTION:Title: Fro
m a greedy Bouncy Particle Sampler to transport accelerated simulated anne
aling\nby Vincent Molin (Chalmers & GU) as part of CAM seminar\n\nLect
ure held in MV:L14.\n\nAbstract\nIn this work we explore the connection be
tween sampling and optimization. Starting from the Bouncy Particle Sampler
\, we combine a continuous time simulated annealing formulation with addit
ional mode-seeking behaviour by biasing the exploration towards modes of t
he objective function. However\, trying to tune the new parameters to achi
eve faster cooling leads us to intead consider a curve of probability meas
ures. The second part of the talk is thus dedicated to considering curve f
ollowing in a space of probability measures. To conclude\, we demonstrate
a particle algorithm inspired by this curve following idea.\n
LOCATION:https://researchseminars.org/talk/cam/49/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Richard Pates (Lund University)
DTSTART:20241021T111500Z
DTEND:20241021T120000Z
DTSTAMP:20260416T215336Z
UID:cam/50
DESCRIPTION:Title: Opt
imal Control on Positive Cones\nby Richard Pates (Lund University) as
part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nIn this talk w
e will discuss a class of optimal control problems for dynamical systems t
hat preserve cones. Under a critical assumption on the cone\, the correspo
nding Bellman equation is satisfied by a linear function\, which can be co
mputed by convex optimization. Three special cases are derived as example
s. The first one\, where the positive cone is the set of positive semi-def
inite matrices\, reduces to standard linear quadratic control. The second
one\, where the positive cone is a polyhedron\, reduces to a recent result
on optimal control of positive systems. The third special case correspond
s to linear quadratic control with additional structure\, such as spatial
invariance\n
LOCATION:https://researchseminars.org/talk/cam/50/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Carlos Jerez Hanckes (Universidad Adolfo Ibañez)
DTSTART:20241022T090000Z
DTEND:20241022T100000Z
DTSTAMP:20260416T215336Z
UID:cam/51
DESCRIPTION:Title: Cel
l Electropermeabilization Modeling Via Multiple Traces Formulation and Tim
e Semi-Implicit Multistep Coupling\nby Carlos Jerez Hanckes (Universid
ad Adolfo Ibañez) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAb
stract\nWe simulate the electrical response of multiple disjoint biologica
l 3D cells undergoing an electropermeabilization process. Instead of solvi
ng the boundary value problem in the unbounded volume\, we reduce it to a
system of boundary integrals equations--the local Multiple Traces Formulat
ion--coupled with nonlinear dynamics on the cell membranes. Though in time
the model is highly non-linear and poorly regular\, the smooth geometry a
llows for boundary unknowns to be spatially approximated by spherical harm
onics. This leads to spectral convergence rates in space. In time\, we use
a multistep semi-implicit scheme. To ensure stability\, the time step nee
ds to be bounded by the smallest characteristic time of the system. Numeri
cal results are provided to validate our claims and future enhancements ar
e pointed out.\n
LOCATION:https://researchseminars.org/talk/cam/51/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Alex (Oleksandr) Misiats (Virginia Commonwealth University)
DTSTART:20241106T090000Z
DTEND:20241106T100000Z
DTSTAMP:20260416T215336Z
UID:cam/52
DESCRIPTION:Title: Reg
ularity and asymptotic properties of nonlocal stochastic evolution equatio
ns arising in chemical and biomedical models\nby Alex (Oleksandr) Mis
iats (Virginia Commonwealth University) as part of CAM seminar\n\nLecture
held in MV:L15.\n\nAbstract\nThis talk is devoted to the influence of stoc
hastic perturbations on the long time behavior of nonlocal evolution equat
ions. I will start with a brief overview of the theory of stochastic evolu
tion equations\, and their applications. I will specifically focus on two
of such models: the bidomain model of heart tissue\, and the aggregation-
diffusion equation (Keller-Segel model).\n\nThe nonlocal character of thes
e equations can be present either in the differential operator (bidomain)
or in the reaction term (Keller-Segel). Using the fundamental concepts in
the area of stochastic analysis\, semigroup theory and PDEs\, in my talk\,
I will address the effects of noise on the existence of global vs. local
solutions (Keller-Segel) and their regularity\, as well as the existence o
f invariant measures (for the bidomain model)\, which is the key step in e
stablishing the qualitative behavior of the underlying physical system. \n
\nThis is a joint project with Prof. Oleksandr Stanzhytskyi from Kyiv Nati
onal University\, Ukraine.\n\nLivestreaming not possible this time due to
unusual room.\n
LOCATION:https://researchseminars.org/talk/cam/52/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Han Zhang (Shanghai Jiao Tong University)
DTSTART:20241122T121500Z
DTEND:20241122T130000Z
DTSTAMP:20260416T215336Z
UID:cam/53
DESCRIPTION:Title: Con
sistent inverse optimal control: methods and applications in robotics\
nby Han Zhang (Shanghai Jiao Tong University) as part of CAM seminar\n\nLe
cture held in MV:L14.\n\nAbstract\nThe inverse linear-quadratic optimal co
ntrol problem is a system identification problem whose aim is to recover t
he quadratic cost function and hence the closed-loop system matrices based
on observations of optimal trajectories. The inverse optimal control fram
ework not only offers a powerful framework to model and predict expert beh
aviors.\n\nIn this talk\, I will introduce a couple of inverse optimal con
trol formulations. Identifiability analysis for each formulation would be
presented. With the persistent excitation assumptions\, we can further sho
w the presented algorithms are statistically consistent. In the end\, inve
rse optimal control algorithms are applied in control algorithm design for
rehabilitation robot.\n
LOCATION:https://researchseminars.org/talk/cam/53/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Eugen Bronasco (University of Geneva)
DTSTART:20250127T121500Z
DTEND:20250127T130000Z
DTSTAMP:20260416T215336Z
UID:cam/54
DESCRIPTION:Title: Alg
ebraic structures and a new order two method for sampling the invariant me
asure of Langevin dynamics\nby Eugen Bronasco (University of Geneva) a
s part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nThis talk ex
plores the problem of sampling the invariant measure of Langevin dynamics
through the lens of exotic aromatic forests — a specialized class of gra
phs forming an algebra that represents the\nalgebra of differential operat
ors. Such algebraic structures play a fundamental role in particular for t
he numerical analysis of numerical integrators. We establish the connectio
n between exotic aromatic forests and numerical methods\, introducing the
relevant algebraic structures and their applications. A key focus is the c
onstruction of a novel method of order two for invariant measure sampling\
, derived using the framework of exotic aromatic forests. Additionally\, w
e demonstrate how these tools\ngenerate order conditions with favorable al
gebraic properties\, facilitate the representation of composition and subs
titution in numerical integrators\, and support backward error analysis an
d the\ncomputation of modified equations. We conclude with an overview of
a new Haskell package designed to automate computations involving graph al
gebras.\n\nPapers and preprints available at: www.unige.ch/~bronasco\n
LOCATION:https://researchseminars.org/talk/cam/54/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Probably no cam
DTSTART:20250407T111500Z
DTEND:20250407T120000Z
DTSTAMP:20260416T215336Z
UID:cam/55
DESCRIPTION:by Probably no cam as part of CAM seminar\n\nLecture held in M
V:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/55/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Eastern
DTSTART:20250421T111500Z
DTEND:20250421T120000Z
DTSTAMP:20260416T215336Z
UID:cam/56
DESCRIPTION:by Eastern as part of CAM seminar\n\nLecture held in MV:L14.\n
Abstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/56/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Probably no cam
DTSTART:20250623T111500Z
DTEND:20250623T120000Z
DTSTAMP:20260416T215336Z
UID:cam/57
DESCRIPTION:by Probably no cam as part of CAM seminar\n\nLecture held in M
V:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/57/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Marie E. Rognes (Simula Research Laboratory)
DTSTART:20250210T121500Z
DTEND:20250210T130000Z
DTSTAMP:20260416T215336Z
UID:cam/58
DESCRIPTION:Title: Bra
in membranes and vasculature: a computational mathematics tale of dimensio
nal gaps\nby Marie E. Rognes (Simula Research Laboratory) as part of C
AM seminar\n\nLecture held in MV:L14.\n\nAbstract\nStructurally\, brain ti
ssue is characterized by thin cell membranes and slender vessels\, definin
g submanifolds of codimension one and two respectively. Functionally\, yo
ur brain fundamentally relies on the transport of ions and nutrients and m
ovement of water in and between these spaces. These physiological processe
s are clearly crucial for brain function and health\, but the precise mech
anisms and their association with neurodegenerative diseases such as Alzhe
imer's and Parkinson's disease and neurological events such as seizures re
main only partially understood. Notably\, mathematical and computational m
odelling are beginning to play an important role in gaining new insight. I
n this talk\, I will discuss key mathematical\, numerical and computationa
l challenges associated with modelling brain mechanics and transport acros
s scales with an emphasis on coupled systems of partial differential equat
ions with dimensional gaps.\n
LOCATION:https://researchseminars.org/talk/cam/58/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Snorre Harald Christiansen (University of Oslo)
DTSTART:20250310T121500Z
DTEND:20250310T130000Z
DTSTAMP:20260416T215336Z
UID:cam/59
DESCRIPTION:Title: Mix
ed finite elements from the point of view of category theory\nby Snorr
e Harald Christiansen (University of Oslo) as part of CAM seminar\n\nLectu
re held in MV:L14.\n\nAbstract\nI will first give an introduction to mixed
finite elements by describing the main example (developed around 1980 by
Raviart\, Thomas and Nédélec but closely related to constructions in dif
ferential topology by de Rham\, Weil and Whitney in the 1940ies). Then I w
ill relate how many fundamental constructions in finite element theory can
be described in terms of category theory. I will not assume prior knowled
ge of the latter. I will argue that the concept of a sheaf provides a unif
ied framework for studying both continuous and discretized PDEs. Sheaf the
ory was first introduced by Leray in the 1940ies and later clarified in th
e Cartan seminar in the 1950ies with in particular algebraic geometry in m
ind. Applied sheaf theory is now a developing subject in many fields.\n
LOCATION:https://researchseminars.org/talk/cam/59/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Olof Runborg (KTH)
DTSTART:20250428T111500Z
DTEND:20250428T120000Z
DTSTAMP:20260416T215336Z
UID:cam/60
DESCRIPTION:Title: Sol
ving Helmholtz Equation Using the Temporal Wave Equation\nby Olof Runb
org (KTH) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nW
e consider iterative methods for the Helmholtz equation that are based on
the related time domain wave equation. In each iteration\, the solution to
the wave equation with a time-periodic forcing is computed and filtered i
n time. For Dirichlet and Neumann problems the iteration corresponds to a
linear and coercive operator which\, after discretization\, is recast as a
positive definite linear system of equations that can be solved with the
conjugate gradient method. The main benefit of using these iterations come
s when considering large scale problems. Implementing parallelized\, high
order methods that run efficiently on big computers is typically easier to
do for time domain methods than for traditional Helmholtz solvers. We als
o show how the iterations can be used to turn a multiscale method for the
wave equation into a multiscale method for the Helmholtz equation.\n
LOCATION:https://researchseminars.org/talk/cam/60/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Balázs Kovács (Paderborn University)
DTSTART:20250317T121500Z
DTEND:20250317T130000Z
DTSTAMP:20260416T215336Z
UID:cam/61
DESCRIPTION:Title: Num
erical surgery for mean curvature flow of surfaces --- A numerical analyst
at the operating table\nby Balázs Kovács (Paderborn University) as
part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nIn this talk I
would like to discuss some numerical algorithms for mean curvature flow o
f two-dimensional surfaces through singularities.\nIt is well known that g
eometric surfaces flows often undergo topological changes. For mean curvat
ure flow this has been heavily investigated by\, e.g.\, Hamilton\, Huisken
and Sinestrari\, and Brendle and Huisken.\nBased on these works we will d
iscuss an algorithm for mean curvature flow with numerical surgery\, a pro
cess which automatically handles these topological changes\, by cutting aw
ay some surface-portions and closing the remaining parts by spherical caps
.\nIn view of this\, the following ethical statement seems appropriate: No
ne of the test subjects were permanently harmed during the experiments.\n
LOCATION:https://researchseminars.org/talk/cam/61/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Larisa Beilina (Chalmers & GU)
DTSTART:20250203T121500Z
DTEND:20250203T130000Z
DTSTAMP:20260416T215336Z
UID:cam/62
DESCRIPTION:Title: Opt
imization approach for solution of parabolic inverse source problem using
partial boundary measurements\nby Larisa Beilina (Chalmers & GU) as pa
rt of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nWe will present
optimization approach for determination of the spatially distributed sour
ce function in the parabolic PDE using partial boundary measurements.\n\nI
n this talk we will:\n1) Present Lagrangian approach for solution of optim
ization problem and derive optimality conditions.\n\n2) Prove that the reg
ularized Tikhnov functional is Frechet differentiable and establish the\ne
xistence and uniqueness of the solution for the inverse problem when\nthe
set of admissible data is bounded.\n\n3) Establish a local stability estim
ate for the unknown source term.\n\n4) Present numerical examples in 2D wi
th noisy data.\n
LOCATION:https://researchseminars.org/talk/cam/62/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Kasper Bågmark (Chalmers & GU)
DTSTART:20250512T111500Z
DTEND:20250512T120000Z
DTSTAMP:20260416T215336Z
UID:cam/63
DESCRIPTION:Title: Rev
isiting nonlinear filtering through deep BSDE methods\nby Kasper Bågm
ark (Chalmers & GU) as part of CAM seminar\n\nLecture held in MV:L14.\n\nA
bstract\nIn this talk\, I present a new approach to the nonlinear filterin
g problem based on the deep Backward SDE (BSDE) method. We begin by formul
ating a system of equations that captures the classical prediction–updat
e steps in Bayesian filtering. While the update step is tractable up to a
normalising constant\, the focus is on approximating the prediction step\,
which involves evolving the prediction density over time. When the hidden
state follows a Stochastic Differential Equation (SDE)\, the prediction d
ensity satisfies the Fokker—Planck equation\, a Partial Differential Equ
ation (PDE). We solve this PDE using a probabilistic BSDE representation\,
which we approximate through an optimisation scheme involving neural netw
orks\, stochastic gradient descent\, and the Euler—Maruyama method. The
approach is demonstrated on numerical examples\, and we numerically examin
e its strong convergence rate.\n
LOCATION:https://researchseminars.org/talk/cam/63/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Johan Ulander (Chalmers & GU)
DTSTART:20250505T111500Z
DTEND:20250505T120000Z
DTSTAMP:20260416T215336Z
UID:cam/64
DESCRIPTION:Title: Bou
ndary-Preserving Weak Approximations of some semilinear Stochastic Partial
Differential Equations\nby Johan Ulander (Chalmers & GU) as part of C
AM seminar\n\nLecture held in MV:L14.\n\nAbstract\nWe propose a boundary-p
reserving numerical scheme for the weak approximations of some scalar-valu
ed stochastic partial differential equations (SPDEs) that only takes value
s in a bounded domain. We only impose regularity assumptions on the drift
and diffusion coefficients locally on the domain. In particular\, the drif
t and diffusion coefficients may be non-globally Lipschitz continuous and
superlinearly growing outside the domain. The scheme consists of a finite
difference discretisation in space and a Lie--Trotter splitting followed b
y exact simulation and exact integration in time. The scheme converges in
the weak sense to the mild solution with rate 1/2 in space and 1/4 in time
for globally Lipschitz continuous test functions. Numerical experiments c
onfirm that the theoretical results are sharp and we compare the proposed
scheme to existing schemes for SPDEs.\n
LOCATION:https://researchseminars.org/talk/cam/64/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Klas Modin (Chalmers & GU)
DTSTART:20250324T121500Z
DTEND:20250324T130000Z
DTSTAMP:20260416T215336Z
UID:cam/65
DESCRIPTION:Title: Emm
y Noether’s theorem\nby Klas Modin (Chalmers & GU) as part of CAM se
minar\n\nLecture held in MV:L14.\n\nAbstract\nI wish to talk about Noether
’s theorem\, including its proof. The subject came up on the occation of
the newly published popular science book “Emmys teorem” by Julia Rava
nis.\n
LOCATION:https://researchseminars.org/talk/cam/65/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Jennifer Ryan (KTH)
DTSTART:20250224T121500Z
DTEND:20250224T130000Z
DTSTAMP:20260416T215336Z
UID:cam/66
DESCRIPTION:Title: Tec
hniques for Data enhancement\nby Jennifer Ryan (KTH) as part of CAM se
minar\n\nLecture held in MV:L14.\n\nAbstract\nExtracting extra information
from data can allow for more insight into the interaction between dispara
te scales. It can also aid in minimising error and decreasing noise in dat
a. While the ability to move data from fine resolutions to coarser resolu
tions is straight forward utilizing a multi-resolution analysis framework\
, moving data from a coarse resolution to a finer resolution while reducin
g errors is more challenging. This relies on combining filtering techniqu
es into the multi-resolution analysis framework. This approach has the fur
ther advantage of requiring fewer computations to gain insight into calcul
ations such as for Bohm speed. In this talk\, we present methods for data
enhancement through multi-resolution analysis and the Smoothness-Increasi
ng Accuracy-Conserving (SIAC) filtering framework. SIAC is known to inhere
ntly take advantage of the underlying physics and allow for the full resol
ution 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.\n
LOCATION:https://researchseminars.org/talk/cam/66/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Simon Jacobsson (KU Leuven)
DTSTART:20250217T121500Z
DTEND:20250217T130000Z
DTSTAMP:20260416T215336Z
UID:cam/67
DESCRIPTION:Title: App
roximating maps into Riemannian manifolds\nby Simon Jacobsson (KU Leuv
en) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nMany in
teresting functions arising in applications map into Riemannian manifolds.
We present an algorithm\, using the manifold exponential and logarithm\,
for approximating such functions. Our approach extends approximation techn
iques for functions into linear spaces in such a way that we can upper bou
nd the forward error in terms of a lower bound on the manifold's sectional
curvature. Furthermore\, when the sectional curvature is nonnegative\, su
ch as for compact Lie groups\, the error is guaranteed to not be worse tha
n in the linear case.\n
LOCATION:https://researchseminars.org/talk/cam/67/
END:VEVENT
BEGIN:VEVENT
SUMMARY:(cancelled) David Krantz (KTH)
DTSTART:20250519T111500Z
DTEND:20250519T120000Z
DTSTAMP:20260416T215336Z
UID:cam/68
DESCRIPTION:by (cancelled) David Krantz (KTH) as part of CAM seminar\n\nLe
cture held in MV:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/68/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Alexander Paul Lewis (Georg-August-University of Goettingen)
DTSTART:20250414T111500Z
DTEND:20250414T120000Z
DTSTAMP:20260416T215336Z
UID:cam/69
DESCRIPTION:Title: Fun
damental theorem for mean square convergence of SDEs on Riemannian manifol
ds\nby Alexander Paul Lewis (Georg-August-University of Goettingen) as
part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nIn recent yea
rs there has been a surge of interest in using stochastic differential equ
ations on Riemannian manifolds for a wide range of purposes. For example i
n: molecular dynamics\, computer vision\, sampling and machine learning. A
s such\, developing accurate algorithms has become critical to estimate so
lutions of manifold valued SDEs in both the weak and strong sense. Moreove
r\, analysis of strong convergence gives us a deeper understanding of the
dynamics and evolution of SDEs\, which has yet to be studied in great deta
il on manifolds.\n\nThe error rate for the Euler(-Murayama) method on Riem
annian manifolds in the weak sense has been established in [1] and was fou
nd to be of global rate 1\; reflecting the classical result known Euclidea
n space. However\, strong convergence rates of the Euler scheme have yet t
o be derived. Though based on intuition\, it is not unreasonable to expect
that the manifold scheme has the same global rate as its Euclidean counte
rpart of 1/2. \n\nBy following closely to the approach laid out in the sem
inal works of Milstein\, we show how to generate high order strong schemes
on a Riemannian manifold with non-positive curvature. In particular\, we
show that the Euler scheme has global rate 1/2\, and we present the Milste
in correction to the Euler scheme which yields a scheme of global order 1.
Finally\, we will formulate the manifold generalisation of the fundamenta
l theorem of strong convergence\, allowing us to obtain global convergence
rates for a wide range of numerical schemes. I will also present numerica
l experiments which illustrate the theoretical guarantees\, as well as a c
ounterintuitive example on 2-dimensional hyperbolic space. \n\nThe talk wi
ll give an overview of results obtained in joint work with Karthik Bharath
and Michael Tretyakov.\n\n\n[1] Bharath\, K.\, Lewis\, A.\, Sharma\, A. a
nd Tretyakov\, M.V.\, 2023. Sampling and estimation on manifolds using the
Langevin diffusion. arXiv preprint arXiv:2312.14882.\n
LOCATION:https://researchseminars.org/talk/cam/69/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Isabel Haasler (Uppsala University)
DTSTART:20250526T111500Z
DTEND:20250526T120000Z
DTSTAMP:20260416T215336Z
UID:cam/70
DESCRIPTION:Title: Sys
tem identification in large populations using optimal transport\nby Is
abel Haasler (Uppsala University) as part of CAM seminar\n\nLecture held i
n MV:L14.\n\nAbstract\nIdentifying the dynamics of individual agents from
aggregate\, population-level data is a key challenge in many applications\
, ranging from biology to crowd dynamics and traffic systems. In recent ye
ars\, optimal transport theory has emerged as a powerful tool for modeling
large populations of agents\, as it bridges a population-level viewpoint
with an agent-level viewpoint.\n\nIn this talk\, I will show how optimal t
ransport can be used to identify agent dynamics from snapshot observations
of the full population. This idea extends even to heterogeneous populatio
ns\, where the observed data results from a superposition of multiple subp
opulations\, each governed by distinct dynamics. In this case\, our method
simultaneously separates the population into ensembles and identifies eac
h ensemble’s governing system\, using only aggregate snapshot observatio
ns.\n
LOCATION:https://researchseminars.org/talk/cam/70/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Souad Mohaoui (Örebro University)
DTSTART:20250331T111500Z
DTEND:20250331T120000Z
DTSTAMP:20260416T215336Z
UID:cam/71
DESCRIPTION:Title: Ten
sor decomposition approaches for motion capture data completion.\nby S
ouad Mohaoui (Örebro University) as part of CAM seminar\n\nLecture held i
n MV:L14.\n\nAbstract\nTensor decompositions are powerful tools for analyz
ing high-dimensional data by breaking down multi-way arrays into smaller\,
meaningful components. They help uncover patterns and handle missing data
effectively. In this work\, we consider two tensor decomposition methods\
, CANDECOMP/PARAFAC (CP) and Tucker\, to address the problem of gap fillin
g in motion capture (MoCap) data. The gap-filling problem in marker-based
MoCap systems occurs when markers become occluded or detached during recor
ding\, resulting in incomplete data. Tensor decompositions offer an effect
ive solution by leveraging the inherent multi-way structure of MoCap data.
We develop and analyze different completion algorithms built upon CP and
Tucker decompositions\, and evaluate their performance across different mi
ssing data scenarios. The algorithms are tested using motion capture seque
nces from the publicly available CMU and HDM05 datasets.\n
LOCATION:https://researchseminars.org/talk/cam/71/
END:VEVENT
BEGIN:VEVENT
SUMMARY:CAM BBQ (tbc)
DTSTART:20250616T111500Z
DTEND:20250616T120000Z
DTSTAMP:20260416T215336Z
UID:cam/72
DESCRIPTION:by CAM BBQ (tbc) as part of CAM seminar\n\nLecture held in MV:
L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/72/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Marc Boxberg (RWTH Aachen University)
DTSTART:20250825T111500Z
DTEND:20250825T120000Z
DTSTAMP:20260416T215336Z
UID:cam/73
DESCRIPTION:Title: Dat
a-Driven Inversion for Ultrasonic Wave Propagation: From Full-Waveform Inv
ersion to Gaussian Process Emulation\nby Marc Boxberg (RWTH Aachen Uni
versity) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nFu
ll-waveform inversion (FWI) is a powerful technique for reconstructing sub
surface properties from wavefield data\, typically employing adjoint-state
methods to compute sensitivity kernels and optimize model parameters. How
ever\, in highly attenuative media\, strong energy dissipation complicates
backpropagation\, posing challenges for gradient-based optimization metho
ds.\nIn this talk\, I will discuss two alternative approaches to wavefield
-based inversion\, using ultrasonic laboratory measurements as an example
application. First\, I will outline a standard FWI approach utilizing conj
ugate-gradient optimization to iteratively refine P-wave velocity and S-wa
ve velocity. Second\, I will present a data-driven alternative: a Gaussian
process emulator that approximates the misfit function over the parameter
space\, enabling efficient inversion without the need for explicit backpr
opagation. This approach offers advantages in scenarios where traditional
FWI struggles due to high attenuation or limited data coverage. However\,
so far\, it has only been tested for determining P- and S-wave velocities
as well as attenuation parameters in homogeneous rock samples. I will disc
uss potential extensions and applications beyond ultrasonic experiments\,
highlighting broader implications for geophysical inversion problems.\n
LOCATION:https://researchseminars.org/talk/cam/73/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Martin Hermann (moved to 24.09\,13:15-14:00) (University of Augsbu
rg)
DTSTART:20250922T111500Z
DTEND:20250922T120000Z
DTSTAMP:20260416T215336Z
UID:cam/74
DESCRIPTION:Title: Rie
mannian optimization methods for ground states of multicomponent Bose-Eins
tein condensates\nby Martin Hermann (moved to 24.09\,13:15-14:00) (Uni
versity of Augsburg) as part of CAM seminar\n\nLecture held in MV:L14.\n\n
Abstract\nGround states of multicomponent Bose-Einstein condensates can be
described as minimizers of the Gross-Pitaevskii energy functional on an i
nfinite-dimensional manifold. For the computation of these minimizers\, we
investigate a family of Riemannian optimization methods with respect to d
ifferent metrics. This allows a unified treatment of several algorithms un
der a common framework and enables us to prove global and local convergenc
e guarantees for important cases. Finally\, we also discuss extensions to
rotating condensates\, where uniqueness can no longer be guaranteed.\n
LOCATION:https://researchseminars.org/talk/cam/74/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Probably no cam
DTSTART:20251013T111500Z
DTEND:20251013T120000Z
DTSTAMP:20260416T215336Z
UID:cam/75
DESCRIPTION:by Probably no cam as part of CAM seminar\n\nLecture held in M
V:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/75/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Probably no cam
DTSTART:20251020T111500Z
DTEND:20251020T120000Z
DTSTAMP:20260416T215336Z
UID:cam/76
DESCRIPTION:by Probably no cam as part of CAM seminar\n\nLecture held in M
V:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/76/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Nikos. I. Kavallaris (Karlstad University)
DTSTART:20250908T111500Z
DTEND:20250908T120000Z
DTSTAMP:20260416T215336Z
UID:cam/77
DESCRIPTION:Title: Net
work Dynamics via Topological Data Analysis and Equation-Free Methods\
nby Nikos. I. Kavallaris (Karlstad University) as part of CAM seminar\n\nL
ecture held in MV:L14.\n\nAbstract\nWe propose a computational framework f
or analyzing the macroscopic dynamics of complex agent-based networks by i
ntegrating Topological Data Analysis (TDA) with the Equation-Free Method.
The approach is demonstrated on Erdős–Rényi random networks. A TDA-bas
ed filtration\, driven by the density of activated nodes\, yields a coarse
macroscopic observable defined via persistent Betti numbers\, enabling si
gnificant dimensionality reduction while preserving essential topology. Wi
thin the Equation-Free framework\, we construct a lifting procedure based
on topological features and identify a data-driven evolution law for this
variable. Finally\, we perform numerical bifurcation and stability analyse
s to characterize the global behavior and qualitative transitions of the e
mergent dynamics. This is a joint work with Konstantinos Spiliotis\, Ole S
önnerborn\, and Haralampos Hatzikirou\n
LOCATION:https://researchseminars.org/talk/cam/77/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Håkon Andreas Hoel (University of Oslo)
DTSTART:20251208T121500Z
DTEND:20251208T130000Z
DTSTAMP:20260416T215336Z
UID:cam/78
DESCRIPTION:Title: Mul
ti-index Monte Carlo method for semilinear stochastic partial differential
equations\nby Håkon Andreas Hoel (University of Oslo) as part of CAM
seminar\n\nLecture held in MV:L14.\n\nAbstract\nWe present an exponential
-integrator based multi-index Monte Carlo method (MIMC) for weak approxima
tions of semilinear stochastic partial differential equations (SPDE). We e
xplore recent theoretical results on multi-index-coupled solutions of SPDE
s\, showing that such couplings are stable and satisfy multiplicative erro
r estimates\, and describe how this theory can be utilized to produce a tr
actable Monte Carlo method for weak approximations. Numerical examples ill
ustrating the performance of MIMC will also be included.\n
LOCATION:https://researchseminars.org/talk/cam/78/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Elisabeth Larsson (Uppsala University)
DTSTART:20251103T121500Z
DTEND:20251103T130000Z
DTSTAMP:20260416T215336Z
UID:cam/79
DESCRIPTION:Title: Num
erical modeling of the human diaphragm\nby Elisabeth Larsson (Uppsala
University) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\
nIn this talk\, we take a look at the challenges involved in modeling the
respiratory mechanics numerically. The focus is on the diaphragm\, which i
s the main respiratory muscle\, and which is very thin with respect to its
size. Two main subproblems\, geometry reconstruction and solution of elas
ticity PDEs in the geometry will be covered. We will discuss the choice of
an unfitted oversampled radial basis function partition of unity method\,
and how it can help us in dealing with the thin geometry. We will discuss
what kind of analysis we can use for this type of method. Finally\, numer
ical results will be shown for linearly elastic model problems in the diap
hragm geometry.\n
LOCATION:https://researchseminars.org/talk/cam/79/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Gerhard Wanner (Université de Genève)
DTSTART:20251110T121500Z
DTEND:20251110T130000Z
DTSTAMP:20260416T215336Z
UID:cam/80
DESCRIPTION:Title: Lan
dmarks in the History of Iterative Methods\nby Gerhard Wanner (Univers
ité de Genève) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbst
ract\nTalk by Gerhard Wanner\n\nJoint work with Martin J. Gander and Phili
ppe Henry\n\nInnehåll\n\nI. Ancient Square Root Calculations (Babylon\, E
gypt\, Heron\, Bakhshali manuscript).\n\nII. Fixed Point Iterations (Jam
shid Ghiath ud-Din Al-Kashi\,\nNityananda\, Habash al-Hasib al-Marwazi\,
J. Kepler\, J. Fourier).\n\nIII. Emergence of Newton's Method (I. Newton\,
Chr. Wren\, J. Wallis and Newton on Kepler's Equation\, Th. Simpson).\n\n
IV. Geometry of Newton's Method (J.~R. Mourraille\, A. Cayley\, G. Julia\,
J. Hubbard\, A. Douady).\n\nV. Error Estimates for Newton's Method (J. Fo
urier\, A.-L. Cauchy\, L.V. Kantorovich).\n
LOCATION:https://researchseminars.org/talk/cam/80/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Michael Roop (Chalmers & GU)
DTSTART:20251215T121500Z
DTEND:20251215T130000Z
DTSTAMP:20260416T215336Z
UID:cam/81
DESCRIPTION:Title: Str
ucture preservation in geophysical fluid dynamics\nby Michael Roop (Ch
almers & GU) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract
\nGeophysical fluid dynamics (GFD) describes flows of fluids on a planetar
y scale. We will discuss the recent contribution to the hierarchy of GFD m
odels developed in [1]\, called thermal quasi-geostrophic (TQG) equations
on the sphere. We will give Arnold’s formulation of TQG. Surprisingly\,
the TQG equations are formulated in terms of Morrison-Greene brackets (sem
idirect product brackets) discovered long ago for another theory - magneto
hydrodynamics (MHD). Further\, we will discuss the structure-preserving di
scretisation for TQG based on Zeitlin’s matrix equations and the Lie-Poi
sson time integrator developed for semidirect products\, and finally we wi
ll see some simulations.\n\n[1] M. Roop\, S. Ephrati. Thermal quasi-geostr
ophic model on the sphere: Derivation and structure-preserving simulation.
Phys. Fluids. 37\, 096601. 2025.\n
LOCATION:https://researchseminars.org/talk/cam/81/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Johan Wärnegård (Chalmers & GU)
DTSTART:20250929T111500Z
DTEND:20250929T120000Z
DTSTAMP:20260416T215336Z
UID:cam/82
DESCRIPTION:Title: Mat
hematical modeling of quantum fluids of light\nby Johan Wärnegård (C
halmers & GU) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstrac
t\nThis talk will revolve around the fascinating topic of quantum fluid of
lights and the mathematical modeling thereof. Quantum fluids of light hav
e attracted a lot of attention lately due to their ability to achieve stro
ng nonlinear effects and to demonstrate macroscopic quantum phenomena such
as superfluidity. In contrast to ultra-cold gases exhibiting superfluidit
y\, these quantum fluids of light constitute open quantum systems as they
continually lose photons to the environment and thus need replenishment. T
his interplay of drive and loss with the fluid-like behavior gives rise to
rich and complex physical phenomena\, which in turn present both challeng
es and opportunities to physicists and applied mathematicians alike.\n
LOCATION:https://researchseminars.org/talk/cam/82/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Sagy Ephrati (Chalmers & GU)
DTSTART:20251027T121500Z
DTEND:20251027T130000Z
DTSTAMP:20260416T215336Z
UID:cam/83
DESCRIPTION:Title: Geo
metric regularisation of two-dimensional turbulence\nby Sagy Ephrati (
Chalmers & GU) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstra
ct\nRegularisation of turbulent flows is essential for computationally eff
icient forecasting. In this talk\, we highlight a geometric mechanism for
regularising two-dimensional turbulence that preserves key conserved quant
ities. We begin with a high-level introduction to how energy is distribute
d across scales in turbulent flows\, using energy spectra to quantify turb
ulence intensity and to illustrate energy cascades that govern flow evolut
ion. We then discuss the conserved quantities unique to two-dimensional tu
rbulence and how they give rise to the inverse energy cascade. This sets t
he stage for recent results on geometric regularisation\, which allows for
control over the energy distribution to enable reduced-complexity computa
tions while retaining fundamental invariants and the inverse energy cascad
e.\n
LOCATION:https://researchseminars.org/talk/cam/83/
END:VEVENT
BEGIN:VEVENT
SUMMARY:CANCELLED Adam Andersson (Saab\, Chalmers & GU)
DTSTART:20251201T121500Z
DTEND:20251201T130000Z
DTSTAMP:20260416T215336Z
UID:cam/84
DESCRIPTION:by CANCELLED Adam Andersson (Saab\, Chalmers & GU) as part of
CAM seminar\n\nLecture held in MV:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/84/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Lucia Swoboda (Chalmers & GU)
DTSTART:20251124T121500Z
DTEND:20251124T130000Z
DTSTAMP:20260416T215336Z
UID:cam/85
DESCRIPTION:Title: Num
erical simulation of beam network models\nby Lucia Swoboda (Chalmers &
GU) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nNetwor
k models are used to describe complex structures in fiber-based materials\
, like paper\, and biological tissues\, such as blood vessels. In these mo
dels\, fibers are represented as beams connected at nodes\, capturing mech
anical properties while reducing computational costs compared to full 3D s
imulations. This talk addresses two problems: (1) the elastic deformation
of beam networks in a stationary setting\, and (2) the elastic wave propag
ation in time-dependent simulations. Both problems\, derived from linear e
lasticity\, describe how materials respond to applied forces and moments.
I will present the key steps in the convergence analysis and error estimat
es\, supported by numerical results. Finally\, I will introduce a two-leve
l overlapping domain decomposition method for efficient parallelization.\n
LOCATION:https://researchseminars.org/talk/cam/85/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Alfred Wärnsäter (KTH Royal Institute of Technology)
DTSTART:20251117T121500Z
DTEND:20251117T130000Z
DTSTAMP:20260416T215336Z
UID:cam/86
DESCRIPTION:Title: Mul
tiple Target Tracking via Multimarginal Optimal Transport\nby Alfred W
ärnsäter (KTH Royal Institute of Technology) as part of CAM seminar\n\nL
ecture held in MV:L14.\n\nAbstract\nThe field of Multiple Target Tracking
(MTT) deals with the task of estimating targets that appear\, disappear\,
and move within a scene\, given data from noisy measurements. An important
step in an MTT algorithm is data association: determining if a detection
is a false positive\, and if not\, which target it should correspond to. W
e find that this task is closely related to the field of multimarginal opt
imal transport (MMOT)\, which originally stems from the problem of optimal
ly moving dirt for the construction of roads. In this work\, we connect MT
T with MMOT. Specifically\, we show how a popular evaluation metric for MT
T (the TGOSPA metric) can be cast as an MMOT problem\, allowing it to be c
omputed efficiently using techniques from the MMOT literature.\n
LOCATION:https://researchseminars.org/talk/cam/86/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Start
DTSTART:20260119T121500Z
DTEND:20260119T130000Z
DTSTAMP:20260416T215336Z
UID:cam/87
DESCRIPTION:by Start as part of CAM seminar\n\nLecture held in MV:L14.\nAb
stract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/87/
END:VEVENT
BEGIN:VEVENT
SUMMARY:End
DTSTART:20260622T111500Z
DTEND:20260622T120000Z
DTSTAMP:20260416T215336Z
UID:cam/88
DESCRIPTION:by End as part of CAM seminar\n\nLecture held in MV:L14.\nAbst
ract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/88/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Mihály Kovács (Pázmány Péter Catholic University and Budapest
University of Technology and Economics and Chalmers University of Technol
ogy)
DTSTART:20260126T121500Z
DTEND:20260126T130000Z
DTSTAMP:20260416T215336Z
UID:cam/89
DESCRIPTION:Title: Neu
mann-Neumann type domain decomposition of elliptic problems on metric grap
hs\nby Mihály Kovács (Pázmány Péter Catholic University and Budap
est University of Technology and Economics and Chalmers University of Tech
nology) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\nWe
develop a Neumann-Neumann type domain decomposition method for elliptic pr
oblems on metric\ngraphs. We describe the iteration in the continuous and
discrete setting and rewrite the latter as a preconditioner\nfor the Schur
complement system. Then we formulate the discrete iteration as an abstrac
t\nadditive Schwarz iteration and prove that it converges to the finite el
ement solution with a rate that is\nindependent of the finite element mesh
size. We also show that the condition number of the Schur complement\nis
bounded uniformly with respect to the finite element mesh size. We discuss
various numerical\nexamples of interest and compare the Neumann-Neumann m
ethod to other preconditioners.\n\nThis research was partially funded by t
he Hungarian National Research\, Development and Innovation\nOffice (NKFIH
) through Grant no. K-145934.\n
LOCATION:https://researchseminars.org/talk/cam/89/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Ana Djurdjevac (University of Oxford)
DTSTART:20260202T121500Z
DTEND:20260202T130000Z
DTSTAMP:20260416T215336Z
UID:cam/90
DESCRIPTION:Title: Mea
n-field optimal control with stochastic leaders\nby Ana Djurdjevac (Un
iversity of Oxford) as part of CAM seminar\n\nLecture held in MV:L14.\n\nA
bstract\nWe study optimal control problems for interacting agent systems a
rysing in opinion dynamics\, where a large number of agents is influenced
by a fixed number of stochastic leaders. We consider a partial mean-field
limit\, leading to a McKean–Vlasov equation for the followers coupled to
controlled stochastic dynamics for the leaders. We show that optimal con
trols for the finite-agent system converge to the optimal control of the l
imiting mean-field system\, providing a low-dimensional and computationall
y efficient approximation of high-dimensional control problems. In additio
n\, we propose efficient numerical methods for computing leader-based cont
rols. We illustrate the theoretical results with numerical experiments for
the Hegselmann–Krause opinion dynamics model. This is joint work with N
. Conrad\, C. Hartmann\, C. Schütte and S. Zimper.\n
LOCATION:https://researchseminars.org/talk/cam/90/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Jennifer K. Ryan (KTH)
DTSTART:20260209T121500Z
DTEND:20260209T130000Z
DTSTAMP:20260416T215336Z
UID:cam/91
DESCRIPTION:Title: Tec
hniques for data enhancement\nby Jennifer K. Ryan (KTH) as part of CAM
seminar\n\nLecture held in MV:L14.\n\nAbstract\nExtracting extra informat
ion from data can allow for more insight into the interaction between disp
arate scales. It can also aid in minimising error and decreasing noise in
data. While the ability to move data from fine resolutions to coarser res
olutions is straight forward utilizing a multi-resolution analysis framewo
rk\, moving data from a coarse resolution to a finer resolution while redu
cing errors is more challenging. This relies on combining filtering techn
iques into the multi-resolution analysis framework. This approach has the
further advantage of requiring fewer computations to gain insight into cal
culations such as for Bohm speed. In this talk\, we present methods for d
ata enhancement through multi-resolution analysis and the Smoothness-Incre
asing Accuracy-Conserving (SIAC) filtering framework. SIAC is known to inh
erently take advantage of the underlying physics and allow for the full re
solution of the approximation and its derivatives in both the physical dom
ain and Fourier signal space. We discuss recent advances and reliance of
the approach on the underlying numerical method that generated the data\n
LOCATION:https://researchseminars.org/talk/cam/91/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Christian Engström (Linnéuniversitetet)
DTSTART:20260309T121500Z
DTEND:20260309T130000Z
DTSTAMP:20260416T215336Z
UID:cam/92
DESCRIPTION:Title: Con
tour integral methods for time-fractional equations\nby Christian Engs
tröm (Linnéuniversitetet) as part of CAM seminar\n\nLecture held in MV:L
14.\n\nAbstract\nWe investigate a class of multiterm time-fractional diffu
sion equations. The problem is reformulated as a perturbed parabolic Volte
rra equation\, which allows the use of Laplace transform techniques and re
solvent estimates. This approach is flexible and covers a wide range of an
omalous diffusion models with combined fractional and classical memory eff
ects. We also discuss implications for the design and analysis of high-ord
er accurate contour integral methods for such equations.\n
LOCATION:https://researchseminars.org/talk/cam/92/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Nicola Guglielmi (Gran Sasso Science Institute)
DTSTART:20260330T111500Z
DTEND:20260330T120000Z
DTSTAMP:20260416T215336Z
UID:cam/93
DESCRIPTION:Title: The
power of rank-1 perturbations in matrix nearness problems.\nby Nicola
Guglielmi (Gran Sasso Science Institute) as part of CAM seminar\n\nLectur
e held in MV:L14.\n\nAbstract\nThis talk concerns matrix nearness problems
related to eigenvalues\, singular values\, and pseudospectra. Such proble
ms arise in a wide range of applications\, including dynamical systems\, w
here they appear in questions of robust stability and control\, as well as
graph-based problems such as clustering and ranking. These problems lead
to algorithms based on structured matrix perturbations that move eigenvalu
es\, singular values\, or Rayleigh quotients to prescribed locations.\n\nR
emarkably\, the optimal perturbations are often of rank one\, or projectio
ns of rank-1 matrices onto a given linear structure\, such as a prescribed
sparsity pattern. In the approach presented here\, these optimal rank-1 p
erturbations are computed by a two-level iteration combining eigenvalue op
timization with root-finding.\n\nThe eigenvalue optimization problem\, wit
h equality or inequality constraints on the perturbation size\, is solved
using gradient-based differential equations for rank-1 matrices. In practi
ce\, this amounts to evolving two vectors that represent the rank-1 matrix
toward a stationary point. The root-finding component determines the opti
mal perturbation size by solving a scalar nonlinear equation. The two algo
rithmic components can either be nested or applied in an alternating fashi
on. Numerical experiments illustrate the efficiency of the approach for se
veral eigenvalue- and singular-value–based nearness problems. \nThe talk
is largely based on a recent monograph written jointly with Christian Lub
ich.\n
LOCATION:https://researchseminars.org/talk/cam/93/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Michael Tretyakov (University of Nottingham)
DTSTART:20260413T111500Z
DTEND:20260413T120000Z
DTSTAMP:20260416T215336Z
UID:cam/94
DESCRIPTION:Title: Neu
ral SDEs for variance reduction\nby Michael Tretyakov (University of N
ottingham) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\n
Variance reduction techniques are of crucial importance for increasing eff
iciency of Monte Carlo simulations. Neural stochastic differential equatio
ns (SDEs)\, with control variates parameterized by neural networks\, are c
onsidered in order to learn approximately optimal control variates and hen
ce reduce variance. A black-box fashion practical variance reduction tool\
, which does not require any lengthy pre-training and tuning\, is proposed
for both SDEs driven by Brownian motion and\, more generally\, by Lévy p
rocesses including those with infinite activity. For the latter case\, opt
imality conditions for the variance reduction are proved. Weak approximati
on of SDEs governed by infinite-activity Lévy processes is also discussed
. Several numerical examples from option pricing are presented. The talk i
s mainly based on a joint work with Piers Hinds.\n
LOCATION:https://researchseminars.org/talk/cam/94/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Lucia Swoboda (Chalmers and GU)
DTSTART:20260504T111500Z
DTEND:20260504T120000Z
DTSTAMP:20260416T215336Z
UID:cam/95
DESCRIPTION:Title: Hyb
ridizable discontinuous Galerkin methods for the wave equation on beam net
work models\nby Lucia Swoboda (Chalmers and GU) as part of CAM seminar
\n\nLecture held in MV:L14.\n\nAbstract\nNetwork models are used to descri
be complex structures found in fiber-based materials such as paper\, or bi
ological tissues. In these models\, fibers are represented as edges connec
ted at nodes\,\nand are modeled as Timoshenko beams to capture the mechani
cal behavior of the material. Understanding wave propagation in such netwo
rks is important for predicting material response and improving performanc
e in applications such as papermaking.\n\nWe present a hybridizable discon
tinuous Galerkin (HDG) method for the spatial discretization of the wave e
quation on fiber networks\, combined with a $\\theta$-scheme for time inte
gration. Through hybridization\, the problem is reformulated as a symmetri
c positive definite system on the network nodes. The HDG spatial discretiz
ation achieves arbitrary-order convergence under mesh refinement without i
ncreasing the size of the global system. We establish convergence and erro
r estimates\, supported by numerical experiments.\n
LOCATION:https://researchseminars.org/talk/cam/95/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Andrii Dmytryshyn (Chalmers and GU)
DTSTART:20260427T111500Z
DTEND:20260427T120000Z
DTSTAMP:20260416T215336Z
UID:cam/96
DESCRIPTION:by Andrii Dmytryshyn (Chalmers and GU) as part of CAM seminar\
n\nLecture held in MV:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/96/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Andreas Myklebust (Norwegian University of Life Sciences)
DTSTART:20260216T121500Z
DTEND:20260216T130000Z
DTSTAMP:20260416T215336Z
UID:cam/97
DESCRIPTION:Title: Fin
ite element discretization of Yang–Mills connections\nby Andreas Myk
lebust (Norwegian University of Life Sciences) as part of CAM seminar\n\nL
ecture held in MV:L14.\n\nAbstract\nIn this talk I will present ongoing wo
rk on finite element discretization of Yang–Mills connections on princip
al bundles. To approximate Yang–Mills connections numerically\, we first
choose local sections and partition the base manifold. This allows us to
represent connections as piecewise 1-forms with trace jumps across interfa
ces prescribed by the sections. For bundles with abelian structure group\,
enforcing these jumps via Lagrange multipliers leads to a well-posed sadd
le-point problem that can be solved using finite element methods. We illus
trate the method on the Hopf bundle.\n
LOCATION:https://researchseminars.org/talk/cam/97/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Adam Andersson
DTSTART:20260525T111500Z
DTEND:20260525T120000Z
DTSTAMP:20260416T215336Z
UID:cam/98
DESCRIPTION:by Adam Andersson as part of CAM seminar\n\nLecture held in MV
:L14.\nAbstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/98/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Easter
DTSTART:20260406T111500Z
DTEND:20260406T120000Z
DTSTAMP:20260416T215336Z
UID:cam/99
DESCRIPTION:by Easter as part of CAM seminar\n\nLecture held in MV:L14.\nA
bstract: TBA\n
LOCATION:https://researchseminars.org/talk/cam/99/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Malin Mosquera (Chalmers and GU)
DTSTART:20260302T121500Z
DTEND:20260302T130000Z
DTSTAMP:20260416T215336Z
UID:cam/100
DESCRIPTION:Title: A
Localised Orthogonal Decomposition Method for Heterogeneous Mixed-Dimensio
nal Problems\nby Malin Mosquera (Chalmers and GU) as part of CAM semin
ar\n\nLecture held in MV:L14.\n\nAbstract\nIn this talk\, we present a mod
el for solving mixed-dimensional elliptic problems with highly heterogeneo
us coefficients\, a type of problem that commonly appears in e.g. modellin
g of fractured porous media but can be computationally challenging to solv
e numerically. Thin structures are modelled as lower-dimensional interface
s embedded in a higher-dimensional bulk domain\, leading to the mixed-dime
nsional model problem.\n\nOur method is based on the Localised Orthogonal
Decomposition (LOD) method and constructs locally supported basis function
s on a coarse mesh that does not resolve the fine-scale variations of the
coefficients. The basis functions are adapted to the problem at hand and t
hus carries the fine-scale information in order to ensure optimal converge
nce with respect to the coarse mesh. This method leads to an exponentially
decaying localisation error. We present numerical experiments to validate
the theoretical findings.\n
LOCATION:https://researchseminars.org/talk/cam/100/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Matthias Geilhufe (Chalmers)
DTSTART:20260323T121500Z
DTEND:20260323T130000Z
DTSTAMP:20260416T215336Z
UID:cam/101
DESCRIPTION:Title: Ul
trafast Thermodynamics - towards entropy\, heat and criticality on picosec
ond timescales\nby Matthias Geilhufe (Chalmers) as part of CAM seminar
\n\nLecture held in MV:L14.\n\nAbstract\nMaterials are composed of a vast
number of ions and electrons\, arranging themselves in regular patterns. D
ue to immense progress in ultrafast spectroscopy\, ranging from the low-fr
equency infrared to high-frequency X-ray regimes\, collective excitations
of charge\, spin\, lattice\, and orbital degrees of freedom can now be tri
ggered and probed on their characteristic time and length scales. Probing
such excitations far from equilibrium motivates the field of ultrafast the
rmodynamics\, which translates well-understood concepts\, such as entropy
and heat\, to picosecond dynamics. I will introduce the concept of ultrafa
st thermodynamics using the example of entropy production due to laser-dri
ven phonons [1] and magnons [2] using stochastic thermodynamics. I will sh
ow that these processes are well described by Langevin equations and intro
duce their derivation on a few examples. Furthermore\, I will show extensi
ons of the formalism to non-Markovian behavior and quantum mechanics [3].\
n\n\n[1] L. Caprini\, H. Löwen & R. M. Geilhufe\, Nature Communications\,
15\, 94 (2024) \n\n[2] F. Tietjen\, R. M. Geilhufe\, PNAS Nexus\, 4\, 3 (
2025)\n\n[3] Y. Qiao\, R. M. Geilhufe\, arXiv:2512.17669\n
LOCATION:https://researchseminars.org/talk/cam/101/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Arnulf Jentzen (School of Data Science & School of Artificial Inte
lligence\, The Chinese University of Hong Kong\, Shenzhen and Institute fo
r Analysis and Numerics\, University of Münster)
DTSTART:20260518T111500Z
DTEND:20260518T120000Z
DTSTAMP:20260416T215336Z
UID:cam/102
DESCRIPTION:Title: Co
mprehensive convergence analysis for the Adam optimizer\nby Arnulf Jen
tzen (School of Data Science & School of Artificial Intelligence\, The Chi
nese University of Hong Kong\, Shenzhen and Institute for Analysis and Num
erics\, University of Münster) as part of CAM seminar\n\nLecture held in
MV:L14.\n\nAbstract\nIn the training of artificial intelligence (AI) syste
ms\, often not the standard gradient descent (GD) method is the employed o
ptimization scheme but instead suitable accelerated and/or adaptive GD met
hods -- such as the momentum and the RMSprop methods -- are considered. Th
e most popular of such accelerated/adaptive optimization methods is presum
ably the Adam optimizer due to Kingma & Ba (2014). In this talk we introdu
ce the notion of the stability region for general deep learning optimizati
on methods and we reveal that among standard GD\, momentum\, RMSprop\, and
Adam we have that Adam is the only optimizer that achieves the optimal hi
gher order convergence speed and also has the maximal stability region. In
another main result of this talk\, which we refer to as Adam symmetry the
orem\, we show for a simple class of quadratic stochastic optimization pro
blems (SOPs) that Adam converges\, as the number of Adam steps increases\,
to the solution of the SOP (the unique minimizer of the strongly convex o
bjective function) if and only if the random variables in the SOP (the dat
a in the SOP) are symmetrically distributed. In particular\, in the standa
rd case where the random variables in the SOP are not symmetrically distri
buted we disprove that Adam converges to the minimizer of the SOP as the n
umber of Adam steps increases. The talk is based on joint works with Steff
en Dereich\, Thang Do\, Robin Graeber\, Sebastian Kassing\, Adrian Riekert
\, and Philippe von Wurstemberger.\n
LOCATION:https://researchseminars.org/talk/cam/102/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Qi Tang (Georgia Tech)
DTSTART:20260601T111500Z
DTEND:20260601T120000Z
DTSTAMP:20260416T215336Z
UID:cam/103
DESCRIPTION:Title: St
ructure-Preserving Neural Operators for Convection–Diffusion Dynamics\nby Qi Tang (Georgia Tech) as part of CAM seminar\n\nLecture held in MV:
L14.\n\nAbstract\nLearning convection–diffusion dynamics with neural ope
rators is difficult because transport and dissipation act on different sca
les\, and standard neural operators often lose stability across regimes. W
e propose a Structure-Preserving Neural Operator that captures this transp
ort–dissipation interplay. The method uses Strang splitting to evolve hy
perbolic and parabolic dynamics in substeps. Convection is handled by a le
arnable semi-Lagrangian approach that follows characteristics and embeds f
low structure directly into the architecture\, while diffusion is treated
through a residual correction neural operator. Experiments on variable-coe
fficient problems and the Vlasov–Poisson–Fokker–Planck system show i
mproved stability\, accuracy\, and long-time performance with large time s
teps.\n
LOCATION:https://researchseminars.org/talk/cam/103/
END:VEVENT
END:VCALENDAR