BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//RLASKEY//CALENDEROUS//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VEVENT
DTSTAMP:20261021T000324Z
LAST-MODIFIED:20160711T154526Z
DTSTART:20160713T190000Z
DTEND:20160713T200000Z
UID:event1620@bu.edu
URL:http://physics.bu.edu/internal/events/show/1620
SUMMARY:Inferring Low-Dimensional Representations of Microstructure Using C
	onvolutional Neural Networks
DESCRIPTION:Featuring Nicholas Lubbers\nPoster: http://physics.bu.edu/inter
	nal//resources/event-image/1620/f60cd08_original\n\nPart of the PhD Final O
	ral Exams.\n\nExamining Committee:  William Klein\, Pankaj Mehta\, Rama Ban
	sil\, Kevin Black\, and Claudio Chamon\n\nAbstract:\n\nMaterials microstruc
	ture effects are crucial to modern high-performance materials. Recent highl
	y successful approaches to materials design use machine learning algorithms
	 for optimization; however\, these approaches have not incorporated microst
	ructure. One requires a faithful\, compact quantification of microstructure
	 to integrate it into the machine learning workflow.\nMotivated by statisti
	cal physics\, we establish a machine learning-driven method that envisions 
	microstructure variations as existing near a low-dimensional manifold. We s
	how how recent work on texture image synthesis using Convolutional Neural N
	etworks (CNNs) can be used to define a faithful representation for microstr
	uctures. This allows us to extract distances between texture images by leve
	raging CNNs that were originally trained for computer vision tasks. We then
	 use manifold learning algorithms to embed these distances into a low-dimen
	sional space\, thereby constructing faithful\, compact microstructure manif
	olds.\n\n![Nick](/resources/event-image/1620/f60cd08_original)
LOCATION:SCI 328\, 590 Commonwealth Avenue\, 02215
STATUS:CONFIRMED
CLASS:PUBLIC
END:VEVENT
END:VCALENDAR