BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Measurement-dependent Noisy Search: An Information Acquisition App
 roach - Prof. Tara Javidi\, University of Calfornia\, San Diego
DTSTART:20180305T143000Z
DTEND:20180305T153000Z
UID:TALK102097@talks.cam.ac.uk
CONTACT:Prof. Ramji Venkataramanan
DESCRIPTION:Information acquisition problems form a class of stochastic co
 ntrol and decision problems in which a decision maker is faced with utiliz
 ing a stochastically varying (and uncontrollable) environment whose state 
 is only partially observable to the decision maker. To best utilize the sy
 stem\, our decision maker can carefully control the acquisition process vi
 a which the set of (noisy) measurements are collected. I will discuss my r
 esearch program to characterize the fundamental (information theoretic) li
 mits and gains associated with dynamically controlling the acquisition pro
 cess.  \n \nIn this talk\, we focus on the special case of measurement-dep
 endent noisy search\; this problem arises in a broad spectrum of applicati
 ons such as medical diagnosis\, spectrum sensing\, sensor management\, ini
 tial access in mmWave communication\, and noisy group testing. We provide 
 a non-asymptotic characterization of the optimal tradeoff between search t
 ime\, resolution\, and search reliability. Our framework rests upon refram
 ing the contributions due to Wald\, Blackwell\, and DeGroot and identifyin
 g the missing link of information acquisition rate in Chernoff’s seminal
  work on active hypothesis testing. Our analysis is sequential in nature a
 nd connects De Groot’s notion of information utility with the Shannon th
 eoretic concept of uncertainty reduction. Our achievability scheme\, i.e. 
 the optimal (adaptive) search strategy\, generalizes the Posterior Matchin
 g of Shayevitz and Feder for channel coding with feedback. \n\nThis work w
 as done in collaborations with my PhD students as well as Y. Kaspi\, O. Sh
 ayevitz\, and M. Wigger. \n\n*Bio*: Tara Javidi studied electrical enginee
 ring at Sharif University of Technology\, Tehran\, Iran from 1992 to 1996.
  She received her MS degrees in electrical engineering (systems) and in ap
 plied mathematics (stochastic analysis) from the University of Michigan\, 
 Ann Arbor\, in 1998 and 1999\, respectively. She received her Ph.D. in ele
 ctrical engineering and computer science from the University of Michigan\,
  Ann Arbor\, in 2002. From 2002 to 2004\, Tara Javidi was an assistant pro
 fessor at the Electrical Engineering Department\, University of Washington
 \, Seattle. In 2005\, she joined the University of California\, San Diego\
 , where she is currently a professor of electrical and computer engineerin
 g.  \n\nTara Javidi was a recipient of the National Science Foundation ear
 ly career award (CAREER) in 2004\, Barbour Graduate Scholarship\, Universi
 ty of Michigan\, in 1999\, and the Presidential and Ministerial Recognitio
 ns for Excellence in the National Entrance Exam\, Iran\, in 1992. Tara Jav
 idi is a Distinguished Lecturer of the IEEE Information Theory Society (20
 17/18).\n
LOCATION:LT6\, Baker Building\, CUED
END:VEVENT
END:VCALENDAR