BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Recent Advances in Structural Learning for Probabilistic Graphical
  Models
DTSTART:20210312T161500
DTEND:20210312T173000
DTSTAMP:20260413T044441Z
UID:c129595f4f0d33833016dadc678b06345cb98bb0cc2ba5016db54e08
CATEGORIES:Conferences - Seminars
DESCRIPTION:Genevera Allen\, Rice University\nProbabilistic Graphical Mo
 dels represent probability distributions as a graph with edges denoting c
 onditional dependence relationships between random variables. These models
  have been studied and developed in computer science\, probability\, and s
 tatistics\, with wide ranging applications in artificial intelligence\, co
 mputer vision\, systems biology\, physics\, and finance\, to name a few.\n
 This talk will review structural learning in probabilistic graphical 
 models\, which seeks to learn the unknown edges or conditional dependencie
 s between random variables\, and highlight new methods and theory from my
  research group for graph structural learning from large and complex da
 ta. Specifically\, I will discuss graph learning for non-Gaussian data
  including data with extreme events\, for mixed data via data integration\
 , for non-simultaneously recorded or non-aligned data via Graph Quilting\,
  and for learning in the presence of latent variables.\nAdditionally\, I
  will present several applications of these approaches to finance\, genomi
 cs\, and neuroscience.  \n\nProbabilistic Graphical Models represent 
 probability distributions as a graph with edges denoting conditional depen
 dence relationships between random variables. These models have been studi
 ed and developed in computer science\, probability\, and statistics\, with
  wide ranging applications in artificial intelligence\, computer vision\, 
 systems biology\, physics\, and finance\, to name a few.\n\nThis talk will
  review structural learning in probabilistic graphical models\, which
  seeks to learn the unknown edges or conditional dependencies between ran
 dom variables\, and highlight new methods and theory from my research grou
 p for graph structural learning from large and complex data. Specifica
 lly\, I will discuss graph learning for non-Gaussian data including dat
 a with extreme events\, for mixed data via data integration\, for non-simu
 ltaneously recorded or non-aligned data via Graph Quilting\, and for lear
 ning in the presence of latent variables.  Additionally\, I will present
  several applications of these approaches to finance\, genomics\, and neur
 oscience.  \n 
LOCATION:https://epfl.zoom.us/j/84927325681
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR