BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Namur Institute For Complex Systems - ECPv6.15.20//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Namur Institute For Complex Systems
X-ORIGINAL-URL:https://www.naxys.be
X-WR-CALDESC:Events for Namur Institute For Complex Systems
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Europe/Brussels
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20230326T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20231029T010000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20240331T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20241027T010000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20250330T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20251026T010000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Europe/Brussels:20240314T130000
DTEND;TZID=Europe/Brussels:20240314T140000
DTSTAMP:20260427T185156
CREATED:20240122T163629Z
LAST-MODIFIED:20240301T100046Z
UID:2088-1710421200-1710424800@www.naxys.be
SUMMARY:Michaël Fanuel (Université de Lille)
DESCRIPTION:Title : Sparsification of the magnetic Laplacian with statistical guarantees \nAbstract : \n  \nLaplacian matrices are discrete analogues of Laplacian differential operators\, which have applications in several areas of applied mathematics such as network science\, machine learning\, control\, etc.\nA Laplacian matrix is in general associated with a graph.\nIn this talk\, we consider the problem of finding a controlled approximation of a Laplacian matrix — called a sparsifier — which only has a few non-zero entries.\nThis (sparse) matrix is associated with a graph with fewer edges\, and is often cheaper to inverse. \nMore specifically\, we consider a U(1)-connection graph\, that is\, a graph where each oriented edge is endowed with a unit modulus complex number which is simply conjugated under orientation flip. A natural replacement for the usual graph Laplacian is then the so-called magnetic Laplacian\, a Hermitian matrix which includes information about the graph’s connection. Connection graphs and magnetic Laplacians appear\, e.g.\, in the problem of angular synchronization (signal processing).In the context of large and dense graphs\, we study spectral sparsifiers of the magnetic Laplacian\, i.e.\, spectral approximations based on subgraphs with few edges. Our approach relies on sampling variants of spanning forests using a custom determinantal point process\, a distribution over edges that favours diversity (and which originates from quantum physics).\nIn a word\, the connected components of our spanning forests are either trees or cycle-rooted trees. The latter partially capture the angular inconsistencies of the connection graph\, and thus provide a way to compress information contained in the connection.\nOne of our contributions is to provide statistical guarantees for a choice of natural estimators of the connection Laplacian based on batches of spanning forests. \nInterestingly\, when the connection graph has weakly inconsistent cycles\, samples of this distribution can be obtained by using a loop-erased random walk.\nThis sampling algorithm is called CyclePopping and is actually rather performant in practice. The law of the number of steps to complete CyclePopping is also known exactly if the connection graph has weakly inconsistent cycles.\nWe shall briefly discuss our contribution in the analysis of this algorithm. \nThis is a joint work with Rémi Bardenet (CNRS & ULille) https://arxiv.org/abs/2208.14797. \n\nThe seminar will take place in Room S08 at the Faculty of Sciences.
URL:https://www.naxys.be/event/michael-fanuel-universite-de-lille/
CATEGORIES:NAXYS Seminar
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Brussels:20240321T130000
DTEND;TZID=Europe/Brussels:20240321T140000
DTSTAMP:20260427T185156
CREATED:20240209T150814Z
LAST-MODIFIED:20240321T095156Z
UID:2098-1711026000-1711029600@www.naxys.be
SUMMARY:Jean-Charles Delvenne (UCLouvain)
DESCRIPTION:Title : Markov chains\, transport theory and statistical physics \nAbstract : \nWe look at the following motivating problem: how to move an electronic memory from a ‘zero’ state to a ‘one’ state\, at minimal energy cost?\n\nMathematically\, this amounts to design a Markov chain  that drives a certain probability measure (encoding a ‘zero’) towards another one (a ‘one’) through an ‘optimal’ path — an avatar of Gaspard Monge’s so-called ‘earth mover problem’\, at the core of transport theory.\n\n We explore various recent results and conjectures around this theme at the interface of statistical physics and Markov chain theory. We support these by illustrations on realistic simulations on electronic memories.\n\nThe seminar will exceptionally take place in Room S06 (and not S08) at the Faculty of Sciences.
URL:https://www.naxys.be/event/jean-charles-delvenne-uclouvain/
CATEGORIES:NAXYS Seminar
END:VEVENT
END:VCALENDAR