BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Synthetic maps for navigating high-dimensional data spaces - Profe
 ssor Alessandro Laio\, SISSA Trieste
DTSTART:20191204T141500Z
DTEND:20191204T151500Z
UID:TALK127330@talks.cam.ac.uk
CONTACT:Lisa Masters
DESCRIPTION:The analysis of large databases aims at obtaining a synthetic 
 description of a system revealing its  salient features. \nWe will describ
 e an approach for charting complex and heterogeneous data spaces\,  provid
 ing a  topography of the high-dimensional  probability distribution from w
 hich the data are harvested.  This topography includes information on the 
 number and the height of the probability peaks\, the depth of the "valleys
 " separating them\, the relative location of the peaks and their hierarchi
 cal organization.  The topography is reconstructed by using an unsupervise
 d variant of Density Peak clustering[Science\, 1492\, vol 322  (2014)] exp
 loiting a non-parametric density estimator[JCTC \,1206\, vol 14 \, (2018) 
 ]\, which automatically measures the density in  the manifold containing t
 he data[Sci Rep. 12140\, vol 7 (2017)]. Importantly\, the density estimato
 r provides an estimate of the error.    This is a key feature\, which allo
 ws distinguishing genuine probability  peaks from density fluctuations due
  to finite sampling.
LOCATION:Department of Chemistry\, Cambridge\, Unilever lecture theatre
END:VEVENT
END:VCALENDAR