RAIN schedule for Autumn Quarter 2014-15
The RAIN seminar is held on Wednesdays from 12:00-1:00pm in Y2E2 101. And yes, lunch is provided!
|September 24||Balasubramanian Sivan||Towards optimal algorithms for prediction with expert advice|
|October 8||Vivek Farias||Online A-B Testing|
|October 29||Fil Menczer||The spread of information in social media|
|December 3||Adam Wierman||TBA
(Joint seminar with OIT and OR)
Google Calendar for RAIN
Previous year's talks
Archived talks can be accessed here.
Talk AbstractsTowards optimal algorithms for prediction with expert advice
Balasubramanian Sivan, Microsoft Research
We study the classic problem of prediction with expert advice. Focusing on settings with a constant number of experts, we develop optimal algorithms and obtain precisely optimal regret values for the case of 2 and 3 experts. Further, we develop regret lower bounds for the multiplicative weights algorithm that exactly match the known upper bounds for an arbitrary number of experts k. This establishes a constant factor separation between the regrets achieved by the optimal algorithm and the multiplicative weights algorithm for a constant number of experts.
The optimal algorithms have a crisp interpretation. For instance, in the case of 2 experts, the optimal algorithm follows the advice of a particular expert with exactly the probability that the expert will turn out to be the best expert.
Our main tool is the minimax principle which lets us analyze the optimal adversary to compute optimal regret values. While analyzing the optimal adversary, we establish deep connections with non-trivial aspects of random walk. We further use this connection to develop an improved regret bound for the case of 4 experts.
Joint work with Nick Gravin and Yuval Peres
Bio: Balasubramanian Sivan is a postdoctoral researcher at Microsoft Research, Redmond. He completed his PhD in Computer Science at the University of Wisconsin-Madison in 2013. His PhD thesis on “Prior Robust Optimization” was awarded the 2014 ACM SIGecom doctoral dissertation award and the University of Wisconsin-Madison Computer Science department’s outstanding graduate student researcher award. His research interests are in Algorithmic Game Theory, online and approximation algorithms, expert learning. More details can be found here: http://research.microsoft.com/en-us/um/people/bsivan/
Online A-B Testing
Vivek Farias, MIT
We consider the problem of A-B testing when the impact of a treatment is marred by a large number of covariates. This is the situation in a number of modern applications of A-B testing (such as in adTech and e-commerce) as well as in more traditional applications (such as clinical trials). Randomization can be highly inefficient in such settings, and thus we consider the problem of optimally allocating test subjects to either treatment with a view to maximizing the efficiency of our estimate of the treatment effect. Our main contribution is a tractable algorithm for this problem in the online setting where subjects arrive, and must be assigned, sequentially. We also characterize the value of optimization and show that it can be expected to grow large with the number of covariates. Finally, using a real-world impression dataset, we show that our algorithms can be expected to yield substantial improvements to efficiency in practice.
Joint work with Nikhil Bhat and Ciamac Moallemi (Columbia GSB)
Bio: Vivek Farias is interested in the development of new methodologies for large scale dynamic optimization and applications in revenue management, finance, marketing and healthcare. He received his Ph.D. in Electrical Engineering from Stanford University in 2007 and has been at MIT since, where he is the Robert N. Noyce Professor of Management. Outside academia, Vivek has contributed to ventures in finance and technology in the role of a researcher, consultant or founder. Vivek is a recipient of an IEEE Region 6 Undergraduate Student Paper Prize (2002), an INFORMS MSOM Student Paper Prize (2006), an MIT Solomon Buchsbaum Award (2008), an INFORMS JFIG paper prize twice (2009, 2011), the NSF CAREER award (2011), MIT Sloan’s Outstanding Teacher award (2013), and was a finalist for the 2011 INFORMS Pierskalla award in healthcare.
The spread of information in social media
Fil Menczer, Yahoo Labs/Indiana University
This talk overviews ongoing work on information diffusion in
social media, focusing in particular on mining, modeling, and
prediction tasks on data from the Twitter network. I present machine
learning efforts that leverage the structure of meme diffusion
networks and many other features to detect misinformation campaigns,
such as astroturf and social bots. We employ modeling techniques to
understand the formation of communities, the creation of social ties,
and the competition for attention. We investigate how different forms
of structural and topical diversity in the network can be leveraged to
predict which memes will go viral. Finally, it time permits, I will
review a few crowdsourcing projects exploring the computation of
unbiased scholarly impact metrics, the anonymous collection of
sensitive data, and social good.
Joint work with many members of the Center for Complex Networks and Systems Research at Indiana University. This research is supported by the National Science Foundation, McDonnell Foundation, and DARPA. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of these funding agencies.
Bio: Filippo Menczer is a visiting scientist at Yahoo Labs in Sunnyvale during 2014-15 while on sabbatical from Indiana University, where he is a professor of informatics and computer science, adjunct professor of physics, and a member of the cognitive science program. He holds a Laurea in Physics from the University of Rome and a Ph.D. in Computer Science and Cognitive Science from the University of California, San Diego. Dr. Menczer has been the recipient of Fulbright, Rotary Foundation, and NATO fellowships, and a Career Award from the National Science Foundation. He serves as director of the Center for Complex Networks and Systems Research and is a Fellow of the Institute for Scientific Interchange Foundation in Torino, Italy, a Senior Research Fellow of The Kinsey Institute, and an ACM Distinguished Scientist. He previously served as division chair in the IUB School of Informatics and Computing, and was Fellow-at-large of the Santa Fe Institute. His research focuses on Web science, social networks, social media, social computation, Web mining, distributed and intelligent Web applications, and modeling of complex information networks. His work has been covered in The New York Times, Wall Street Journal, NPR, CNN, The Economist, Nature, Science, The Washington Post, Wired, The Atlantic, BBC, and many other US and international news sources.