BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Screening strategies in the presence of interactions - Dean\, A (O
 hio State)
DTSTART:20110902T130000Z
DTEND:20110902T133000Z
UID:TALK32619@talks.cam.ac.uk
CONTACT:Mustapha Amrani
DESCRIPTION:Screening is the process of using designed experiments and sta
 tistical analyses to search through a large number of potentially influent
 ial factors in order to discover the few factors that have a substantial e
 ffect on a measured response (i.e. that are "active"). In this setting\, c
 onventional fractional factorial experiments typically require too many ob
 servations to be economically viable. To overcome this problem in practice
 \, interactions are often dropped from consideration and assumed to be neg
 ligible\, sometimes without substantive justification. Such loss of inform
 ation can be a serious problem in industrial experimentation because explo
 itation of interactions is a key tool for product improvement. This talk d
 escribes an assessment and comparison of two screening strategies for inte
 ractions\, namely supersaturated designs and group screening\, together wi
 th a variety of data analysis methods\, based on shrinkage regression and 
 Bayesian methods. Recommendation s on the use of the screening strategies 
 are provided through simulation studies. \n
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR