BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Hawkes process as models for some genomic data - Patricia Reynaud-
 Bouret (CNRS and Univ. Nice)
DTSTART:20110520T150000Z
DTEND:20110520T160000Z
UID:TALK30797@talks.cam.ac.uk
CONTACT:Richard Nickl
DESCRIPTION:It seems that some of the genomic data\, such as positions of 
 words on\n the DNA or\n positions of transcription regulatory elements may
  hint for synergies\n between\n them. One of the statistical possible mode
 l to catch those\n interactions is the\n Hawkes process\, which has been f
 irst introduced to model earthquakes.\n Gusto and\n Schbath have introduce
 d this model for genome analysis.  However if\n maximum\n likelihood metho
 ds exist and if AIC criterion is usually used to\n select a\n correct numb
 er of parameters\, this combination has been proved to be\nnot\naccurate w
 hen the complexity of the family of parametric models is\nhigh. After\ndis
 cussing the Hawkes models (multivariate or not) and explaining what\nhas b
 een\n done from a parametric point of view (eventually combined with AIC)\
 , I\nwill\nexplain what adaptive model selection can and cannot do and als
 o what\nthresholding in certain cases and Lasso methods may improve.\n\nht
 tp://math.unice.fr/~reynaudb/
LOCATION:MR12\, CMS\, Wilberforce Road\, Cambridge\, CB3 0WB
END:VEVENT
END:VCALENDAR