BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//RLASKEY//CALENDEROUS//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20260828T113317Z
LAST-MODIFIED:20160311T142644Z
DTSTART:20160314T180000Z
DTEND:20160314T190000Z
UID:event1571@bu.edu
URL:http://physics.bu.edu/internal/events/show/1571
SUMMARY:New Perspectives on High Dimensional Inference
DESCRIPTION:Featuring Madhu Advani\, Stanford University\n\nPart of the Bio
	physics/Condensed Matter Seminar Series.\n\nTo model modern large-scale dat
	asets\, we need efficient algorithms to infer a set of P unknown model para
	meters from N noisy measurements. What are fundamental limits on the accura
	cy of parameter inference\, given limited measurements\, signal-to-noise ra
	tios\, prior information\, and computational tractability requirements? How
	 can we combine prior information with measurements to achieve these limits
	? Classical statistics gives incisive answers to these questions as the mea
	surement density approaches infinity. However\, modern 'big data' problems 
	are often high-dimensional: have finite measurement density (N/P). This reg
	ime is important for a variety of fields and to study it we formulate and a
	nalyze high-dimensional inference as a problem in the statistical physics o
	f quenched disorder. This analysis reveals that widely cherished Bayesian i
	nference algorithms are suboptimal\, and yields tractable\, optimal algorit
	hms to replace them.
LOCATION:SCI 328\, 590 Commonwealth Avenue\, 02215
STATUS:CONFIRMED
CLASS:PUBLIC
END:VEVENT
END:VCALENDAR