BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:The Strange Case of Privacy in Equilibrium Models - Mallesh Pai (R
 ice University\; University of Pennsylvania )
DTSTART:20161028T150000Z
DTEND:20161028T160000Z
UID:TALK68659@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:Joint work with Rachel Cummings\, Katrina Ligett and Aaron Rot
 h<br><span><br>The literature on differential privacy by and large takes t
 he data set being being analyzed as exogenously given. As a result\, by va
 rying a privacy parameter in his algorithm\, the analyst straightforwardly
  chooses the potential privacy loss of any single entry in the data set.&n
 bsp\; Motivated by privacy concerns on the internet\, we consider a styliz
 ed setting where the dataset is endogenously generated\, depending on the 
 privacy parameter chosen by the analyst. In our model\, an agent chooses w
 hether to purchase a product. This purchase decision is recorded\, and a d
 ifferentially private version of his purchase decision may be used by an a
 dvertiser to target the consumer. A change in the privacy parameter theref
 ore affects\, in equilibrium\, the agents&#39\; purchase decision\, the pr
 ice of the product\, and the targeting rule used by the advertiser. We dem
 onstrate that the comparative statics with respect to privacy parameter ma
 y be exactly reversed relative to the exogenous data set benchmark\, for e
 xample a higher privacy parameter may nevertheless be more informative etc
 .&nbsp\; More care is needed in understanding the effects of private analy
 sis of a data set that is endogenously generated.</span>
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR