BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Modelling Network Data - Patrick Wolfe\, University College London
DTSTART:20120525T150000Z
DTEND:20120525T160000Z
UID:TALK36572@talks.cam.ac.uk
CONTACT:Richard Samworth
DESCRIPTION:Networks are fast becoming a primary object of interest in sta
 tistical\ndata analysis\, with important applications spanning the social\
 ,\nbiological\, and information sciences.  A common aim across these\nfiel
 ds is to test for and explain the presence of structure in network\ndata. 
 In this talk we show how characterizing the structural features\nof a netw
 ork corresponds to estimating the parameters of various\nrandom network mo
 dels\, allowing us to obtain new results for\nlikelihood-based inference a
 nd uncertainty quantification in this\ncontext.  We discuss asymptotics fo
 r stochastic blockmodels with\ngrowing numbers of classes\, the determinat
 ion of confidence sets for\nnetwork structure\, and a more general point p
 rocess modeling for\nnetwork data taking the form of repeated interactions
  between senders\nand receivers\, where we show consistency and asymptotic
  normality of\npartial-likelihood-based estimators related to the Cox prop
 ortional\nhazards model (arXiv:1201.5871\, 1105.6245\, 1011.4644\, 1011.17
 03).\n
LOCATION:MR12\, CMS\, Wilberforce Road\, Cambridge\, CB3 0WB
END:VEVENT
END:VCALENDAR