BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Utilizing Gaussian Boson Sampling for Hyperspectral Data Reduction
DTSTART;VALUE=DATE-TIME:20260911T070000Z
DTEND;VALUE=DATE-TIME:20260911T074000Z
DTSTAMP;VALUE=DATE-TIME:20260922T173908Z
UID:indico-contribution-423-2048@events.ncbj.gov.pl
DESCRIPTION:Speakers: Jakub Mielczarek (Jagiellonian University)\nGaussian
  Boson Sampling (GBS) is a photonic sampling process whose output probabil
 ities can be related to graph-theoretic problems\, including clique findin
 g. In this talk\, I will discuss the application of GBS to the band-select
 ion problem in hyperspectral satellite data\, where the goal is to identif
 y a reduced set of spectral bands while preserving the relevant informatio
 n. \nAlthough GBS was originally formulated as a photonic process\, it is 
 also possible to explore its implementation on qubit-based quantum archite
 ctures\, which are currently more widely accessible. I will present our re
 cent construction that enables such a mapping. Finally\, I will discuss th
 e advantages and limitations of employing GBS to satellite data analysis.\
 n\nhttps://events.ncbj.gov.pl/event/468/contributions/2048/
LOCATION:
URL:https://events.ncbj.gov.pl/event/468/contributions/2048/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Maximum likelihood using multidimensional hyperplane sections
DTSTART;VALUE=DATE-TIME:20260911T074000Z
DTEND;VALUE=DATE-TIME:20260911T081000Z
DTSTAMP;VALUE=DATE-TIME:20260922T173908Z
UID:indico-contribution-423-2033@events.ncbj.gov.pl
DESCRIPTION:Speakers: Kacper Topolnicki (Jagiellonian University)\nA more 
 general version of the “plane section” algorithm from [Computer Physic
 s Communications\, 319:109913\, Feb. 2026] is used to represent probabilit
 y distributions for likelihood optimization. The algorithm from [Computer 
 Physics Communications\, 319:109913\, Feb. 2026] was implemented in *pytho
 n* using the *pytorch* library allowing gradients to be calculated with re
 spect to the parameters of the representation. This in turn made it possib
 le to perform maximum likelihood estimations using gradient methods\, for 
 example the *Adam* optimizer. Applications of the previous implementation 
 were limited to two dimensions. The new version\, described in the talk\, 
 can in principle be used with an arbitrarily large number of dimensions. T
 his number is of course subjected to a limit determined by the available c
 omputing resources. The new algorithm is also much simpler conceptually op
 ening the possibility of creating implementations in other programming lan
 guages. Additionally\, the proposed algorithm might allow the analytical c
 alculation of gradients for use in custom automatic differentiation functi
 ons. The implementation is tested using two and three dimensional distribu
 tions related to the problem of finding charged particle tracks in gas det
 ectors of high energy physics experiments.\n\nhttps://events.ncbj.gov.pl/e
 vent/468/contributions/2033/
LOCATION:
URL:https://events.ncbj.gov.pl/event/468/contributions/2033/
END:VEVENT
END:VCALENDAR