BEGIN:VCALENDAR
PRODID:-//eluceo/ical//2.0/EN
VERSION:2.0
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:11de3d8523cae9cf36f202e7372b4b65
DTSTAMP:20260505T174019Z
SUMMARY:CCMI Seminars – Rima Alaifari (RWTH Aachen University)
DESCRIPTION:Speaker: Rima Alaifari (RWTH Aachen University)\nTitle: Recover
 ing and learning functions from representations\nAbstract: Recovering and 
 learning functions from partial or indirect observations is a fundamental 
 challenge in many applications. This talk explores two distinct but concep
 tually related problems: phase retrieval from Gabor and wavelet transform 
 magnitudes\, and nonparametric regression via signature transforms.\nPhase
  retrieval seeks to reconstruct a function from the magnitude of its linea
 r transform\, raising questions of uniqueness and stability. In contrast\,
  nonparametric regression aims to estimate statistical relationships in hi
 gh-dimensional time-series data using nonlinear hierarchical representatio
 ns. While these problems arise in different contexts\, both rely on struct
 ured signal representations—the Gabor transform and wavelets providing l
 inear feature hierarchies\, while the signature transform captures nonline
 ar dependencies in sequential data. I will discuss key aspects of these re
 presentations and their implications for function recovery and estimation.
URL:https://www.imperial.ac.uk/events/206135/ccmi-seminars-rima-alaifari-rw
 th-aachen-university/
DTSTART;TZID=Europe/London:20260224T140000
DTEND;TZID=Europe/London:20260224T150000
LOCATION:CHEM 637\, Chemistry\, South Kensington Campus\, Imperial College 
 London\, London\, SW7 2AZ\, United Kingdom
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:STANDARD
DTSTART:20260224T140000
TZNAME:GMT
TZOFFSETTO:+0000
TZOFFSETFROM:+0000
END:STANDARD
END:VTIMEZONE
END:VCALENDAR