- talks - lists.cs.princeton.edu

ORFE Talk: The Implications of Privacy-Aware Choice, Today, 4:30pm
by Nicole E. Wagenblast 07 Feb '17

07 Feb '17

ORFE Colloquium Speaker Rachel Cummings, from California Institute of Technology Tuesday, February 7th- 4:30pm Sherrerd Hall, Room 101 The Implications of Privacy-Aware Choice Privacy concerns are becoming a major obstacle to using data in the way that we want. It's often unclear how current regulations should translate into technology, and the changing legal landscape surrounding privacy can cause valuable data to go unused. In addition, when people know that their current choices may have future consequences, they might modify their behavior to ensure that their data reveal less --- or perhaps, more favorable --- information about themselves. Given these concerns, how can we continue to make use of potentially sensitive data, while providing satisfactory privacy guarantees to the people whose data we are using? Answering this question requires an understanding of how people reason about their privacy and how privacy concerns affect behavior. In this talk, we will see how strategic and human aspects of privacy interact with existing tools for data collection and analysis. I will begin by adapting the standard model of consumer choice theory to a setting where consumers are aware of, and have preferences over, the information revealed by their choices. In this model of privacy-aware choice, I will show that little can be inferred about a consumer's preferences once we introduce the possibility that she has concerns about privacy, even when her preferences are assumed to satisfy relatively strong structural properties. Next, I will analyze how privacy technologies affect behavior in a simple economic model of data-driven decision making. Intuition suggests that strengthening privacy protections will both increase utility for the individuals providing data and decrease usefulness of the computation. I will demonstrate that this intuition can fail when strategic concerns affect consumer behavior. Finally, I'll discuss ongoing behavioral experiments, designed to empirically measure how people trade off privacy for money, and to test whether human behavior is consistent with theoretical models for the value of privacy.

1 0

ORFE Talk: The Implications of Privacy-Aware Choice, Tuesday, Feb 7, 4:30pm- Reminder
by Nicole E. Wagenblast 06 Feb '17

06 Feb '17

ORFE Colloquium Speaker Rachel Cummings, from California Institute of Technology Tuesday, February 7th- 4:30pm Sherrerd Hall, Room 101 The Implications of Privacy-Aware Choice Privacy concerns are becoming a major obstacle to using data in the way that we want. It's often unclear how current regulations should translate into technology, and the changing legal landscape surrounding privacy can cause valuable data to go unused. In addition, when people know that their current choices may have future consequences, they might modify their behavior to ensure that their data reveal less --- or perhaps, more favorable --- information about themselves. Given these concerns, how can we continue to make use of potentially sensitive data, while providing satisfactory privacy guarantees to the people whose data we are using? Answering this question requires an understanding of how people reason about their privacy and how privacy concerns affect behavior. In this talk, we will see how strategic and human aspects of privacy interact with existing tools for data collection and analysis. I will begin by adapting the standard model of consumer choice theory to a setting where consumers are aware of, and have preferences over, the information revealed by their choices. In this model of privacy-aware choice, I will show that little can be inferred about a consumer's preferences once we introduce the possibility that she has concerns about privacy, even when her preferences are assumed to satisfy relatively strong structural properties. Next, I will analyze how privacy technologies affect behavior in a simple economic model of data-driven decision making. Intuition suggests that strengthening privacy protections will both increase utility for the individuals providing data and decrease usefulness of the computation. I will demonstrate that this intuition can fail when strategic concerns affect consumer behavior. Finally, I'll discuss ongoing behavioral experiments, designed to empirically measure how people trade off privacy for money, and to test whether human behavior is consistent with theoretical models for the value of privacy.

1 0

Shilpa Nadimpalli will present her pre-FPO for on Monday, February 13, 2017 at 9am in CS 401.
by Nicki Gotsis 03 Feb '17

03 Feb '17

Shilpa Nadimpalli will present her pre-FPO for on Monday, February 13, 2017 at 9am in CS 401. The members of her committee are as follows: Mona Singh (Adviser); Readers: Ben Raphael and Olga Troyanskaya; Nonreaders: Barbara Engelhardt and Tom Funkhouser The talk abstract follows below: Proteins play essential roles in nearly every biological process, from metabolism and sensory perception to immune response and gene regulation. The vast majority of these functions are carried out through specific protein interactions with DNA, metabolites and other ligands, forming complex biological networks. Although mutations affecting protein interactions have resulted in human disease, other protein interactions evolve naturally within the human population and across species. Experimental techniques to determine the prevalence and functional impact of mutations affecting protein interactions are prohibitively costly and time-consuming (e.g., x-ray crystallography, protein binding microarrays). In my thesis, I develop and use novel computational approaches to measure the extent to which protein interactions are perturbed across healthy individuals, across individuals with disease, and across species. In my talk, I will first describe a tool I have developed to automatically infer interaction sites from sequence data. This tool can infer interaction sites in over 60% of human genes---a nearly threefold increase in coverage over related state-of-the-art methods. Next, I will focus on utilizing this tool to identify genes whose interaction sites are enriched in cancer mutations by contrasting observed natural variants from tens of thousands of healthy individuals to observed cancer mutations from thousands of tumor samples. These genes, often rarely mutated across a cohort of patients and thus easily missed by more standard approaches, are likely to be critically relevant for cancer development. In the last part of my talk, I will discuss my application of a comparative genomics framework to quantify changes in protein-DNA interactions across species. I find substantial evidence suggesting that, contrary to popular convention, gene regulatory network perturbation can stem from changes in non-duplicated DNA-binding proteins. --

1 0

ORFE Talk: The Implications of Privacy-Aware Choice, Tuesday, Feb 7, 4:30pm
by Nicole E. Wagenblast 01 Feb '17

01 Feb '17

ORFE Colloquium Speaker Rachel Cummings, from California Institute of Technology Tuesday, February 7th- 4:30pm Sherrerd Hall, Room 101 The Implications of Privacy-Aware Choice Privacy concerns are becoming a major obstacle to using data in the way that we want. It's often unclear how current regulations should translate into technology, and the changing legal landscape surrounding privacy can cause valuable data to go unused. In addition, when people know that their current choices may have future consequences, they might modify their behavior to ensure that their data reveal less --- or perhaps, more favorable --- information about themselves. Given these concerns, how can we continue to make use of potentially sensitive data, while providing satisfactory privacy guarantees to the people whose data we are using? Answering this question requires an understanding of how people reason about their privacy and how privacy concerns affect behavior. In this talk, we will see how strategic and human aspects of privacy interact with existing tools for data collection and analysis. I will begin by adapting the standard model of consumer choice theory to a setting where consumers are aware of, and have preferences over, the information revealed by their choices. In this model of privacy-aware choice, I will show that little can be inferred about a consumer's preferences once we introduce the possibility that she has concerns about privacy, even when her preferences are assumed to satisfy relatively strong structural properties. Next, I will analyze how privacy technologies affect behavior in a simple economic model of data-driven decision making. Intuition suggests that strengthening privacy protections will both increase utility for the individuals providing data and decrease usefulness of the computation. I will demonstrate that this intuition can fail when strategic concerns affect consumer behavior. Finally, I'll discuss ongoing behavioral experiments, designed to empirically measure how people trade off privacy for money, and to test whether human behavior is consistent with theoretical models for the value of privacy.

1 0

EE Department Colloquium - Korhammer Lecture Series
by Nicole E. Wagenblast 30 Jan '17

30 Jan '17

Having trouble viewing this email? Click here Korhammer Lecture Series The Science of Information: Case Studies from DNA and RNA Assembly David Tse Professor, Information Systems Laboratory, Department of Electrical Engineering Stanford University February 22, 2017 5:00 pm - 6:00 pm Bowen Hall Auditorium Reception to follow Click here for more information This seminar is supported by the Korhammer Lecture Fund Copyright © 20XX. All Rights Reserved. Princeton University, Department of Electrical Engineering , Olden Street , Princeton , NJ 08544 SafeUnsubscribe™ heathere(a)princeton.edu Forward this email | Update Profile | About our service provider Sent by heathere(a)princeton.edu in collaboration with Try it free today Save

1 0

Aaron Blankstein will present his Pre FPO on Tuesday, February 7, 2017 at 10am in CS 302.
by Nicki Gotsis 27 Jan '17

27 Jan '17

Aaron Blankstein will present his Pre FPO on Tuesday, February 7, 2017 at 10am in CS 302. The members of his committee are as follows: Readers: Michael Freedman (adviser), Jennifer Rexford, and Nick Feamster; Non-Readers: Robert Tarjan and Kyle Jamieson Everyone is invited to attend his talk. The talk abstract follows below. Abstract: In my thesis, I will show how, in many respects, modern web applications are built with implicit structure, such that several classic problems in computer systems can be tackled in new ways allowing applications to reap security and performance benefits. I will apply this thesis in two separate contexts. In the first, I will look at the problems of providing some automatic security to web applications by partitioning the server-side code of the application, isolating those partitions, learning access control policies for those partitions, and extending this isolation to front-end code with javascript sandboxing. In the second, I will look at how modern web applications interact with application caches, and develop a family of cache eviction policies tailored to these needs by focusing on prioritizing individual items to capture how item request frequencies, associated costs, and expiration times affect cache performance.

1 0

Seminar Reminder: Physical Knowledge Discovery in the Internet of Things. Fri. Jan 20 noon-1:30 in CS 402.
by Margaret Martonosi 20 Jan '17

20 Jan '17

Physical Knowledge Discovery in the Internet of Things Prof. Pei Zhang Carnegie Mellon University Friday Jan 20 noon-1:30 in CS 402 Abstract: In many sensing scenarios, direct sensing of desired events is difficult and often impossible. This is often due to deployment difficulties, lack of available sensors, and cost of maintenance. This talk will explore indirect sensing framework that infer information from physical inertial sensing through a combination of empirical, analytical and heuristic models. The talk then explore this framework through three projects. 1) The SensorFly system, a low-cost, miniature aerial sensor network that aims to be autonomous in deployment, maintenance and adaptation to the environment. By modeling the motion and movement, these sensors collaboratively localize and navigate indoors using inertial sensors. 2) Landing the sensors inside the building, the same sensor sets can be used to discover occupant information including location, identity, status, etc. 3) The same sensors on the human body, the system can inference muscle activity and fatigue level. The talk will explore the commonalities of data understanding and inferencing. Bio: Pei Zhang is an associate research professor in the ECE departments at Carnegie Mellon University. He received his bachelor's degree with honors from California Institute of Technology in 2002, and his Ph.D. degree in Electrical Engineering from Princeton University in 2008. While at Princeton University, he developed the ZebraNet system, which is used to track zebras in Kenya. It was the first deployed, wireless, ad- hoc, mobile sensor network. His recent work includes SensorFly (focus on groups of autonomous miniature-helicopter based sensor nodes) and MARS (Muscle Activity Recognition). Beyond research publications, his work has been featured on popular media including CNN, Science Channel, Discovery Channel, CBS News, CNET, Popular Science, BBC Focus, etc. He is also a cofounder of the startup Vibradotech. In addition, he has won several awards including the NSF CAREER, Edith and Martin B. Stein Solar Energy Innovation Award, and a member of the Department of Defense Computer Science Studies Panel. Margaret Martonosi H. T. Adams '35 Professor of Computer Science. Princeton University ■ mrm(a)princeton.edu ■ http://www.princeton.edu/~mrm

1 1

Reid Oda will present his Pre FPO January 24, 2017 in CS 402 at 12pm.
by Nicki Gotsis 18 Jan '17

18 Jan '17

Reid Oda will present his Pre FPO January 24, 2017 in CS 402 at 12pm. The members of his committee are: Rebecca Fiebrink (adviser), Adam Finkelstein, Dan Trueman (readers), Chris Chafe (Stanford), Marshini Chetty (non-readers) Everyone is invited to attend his talk. The talk abstract follow below: I will consider the problem of playing live, rhythmically synchronized, collaborative music over the Internet. Though the Internet has revolutionized many forms of communication, online collaborative musical performance has yet to become widespread. This is largely because of the latency (temporal delay) that is inherent in Internet communication, which makes it difficult for musicians to play in time with one another. In my thesis I propose two solutions to this problem, deploy them to musicians, and study them. The first approach, called the Global Metronome, allows players to play in time with one another, even if there is zero communication between them. The Global Metronome is a GPS-connected tempo-keeping device that automatically beats in time with every other Global Metronome implementation. The approach also allows for MIDI (musical instrument digital interface) clock enabled devices to be synchronized over long distances, which has not been previously possible. The second approach, MalLo, is a predictive instrument used to combat latency via prediction. By tracking a percussion mallet, MalLo predicts when a note is going to be played, and sends the timing information across a network, early, so that it can be turned into audio at the same time by both sender and receiver. Finally, I will provide a comprehensive analysis of the strengths and weaknesses of each approach and make recommendations for further advancements.

1 0

Nick Giannarakis will present his general exam on January 24, 2017 at 9:30am in CS 302.
by Nicki Gotsis 18 Jan '17

18 Jan '17

Nick Giannarakis will present his general exam on January 24, 2017 at 9:30am in CS 302. The members of his committee are: Andrew Appel (adviser), Margaret Martonosi, and David Walker. Everyone is invited to attend his talk, and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below. Title: Semantics of well synchronized programs on weakly consistent machines. Abstract: The semantics of shared-memory concurrent programs are quite intricate. For one, optimizing compilers perform transformations that may be unsound with respect to concurrent programs. Moreover, modern multiprocessors have relaxed semantics: the widespread assumption of sequential consistency is invalidated by common architectural optimizations. Processors reorder instructions (or executes them out of order), and techniques such as caching of memory data and buffering of operations break the abstraction of a single global view of the memory shared by all threads. Programs that don’t have data races, however, should execute as if they were running on a sequentially consistent machine. In this work, we focus on a class of programs called well synchronized; access to some part of the shared memory is arbitrated by synchronizing operations and threads do not compete with each other for that memory. Using the Coq proof assistant, we formalized this notion of well synchronized programs, and used it to prove that cooperative (yield on synchronizing operations) and preemptive (interleaving) semantics of such programs coincide. Furthermore, we proved that well synchronized programs don’t have data races, hence their relaxed semantics coincides with their interleaving semantics. We prove that result in a generic and modular way, and demonstrate its applicability on the TSO memory model using the X86 ISA. Reading List: • Stability in weak memory models, Jade Alglave, Luc Maranget. In CAV 2010. • Relaxed Memory Concurrency and Verified Compilation. Jaroslav Sevcik, Viktor Vafeiadis, Francesco Zappa Nardelli, Suresh Jagannathan, Peter Sewell. In POPL 2011. • Reasoning about the Implementation of Concurrency Abstractions on x86-TSO, Scott Owens. In ECOOP 2010. • Compositional CompCert. Gordon Stewart, Lennart Beringer, Santiago Cuellar, Andrew W. Appel. In POPL 2015. • Taming Weak memory models.Sizhuo Zhang, Arvind, Muralidaran Vijayaraghavan. Draft. • Verifying read-copy-update in a logic for weak memory. Joseph Tassarotti, Derek Dreyer, Viktor Vafeiadis. In PLDI 2015. • A promising semantics for relaxed-memory concurrency. Jeehoon Kang, Chung-Kil Hur, Ori Lahav, Viktor Vafeiadis, Derek Dreyer. In POPL 2017. • Types and Programming Languages, Benjamin C. Pierce. • The Formal Semantics of Programming Languages, Glynn Winskel.

1 0

Prof. Pei Zhang. CMU. Friday Seminar Jan 20. Noon-1:30.
by Margaret Martonosi 13 Jan '17

13 Jan '17

HI folks- Pei Zhang from CMU Silicon Valley will be visiting on Friday Jan 20 and giving a seminar. During his PhD, Pei was a lead student on the Princeton ZebraNet project, and has since continued to do some very cool work in the area of mobile sensing and IoT. The seminar will be Friday Jan 20 noon-1:30 in CS 402 and I’ll provide lunch. The title/abstract/bio are below. If you are interested in having a spot on his one-on-one schedule, please let me know! Thanks very much, Margaret Title: Physical Knowledge Discovery in the Internet of Things Abstract: In many sensing scenarios, direct sensing of desired events is difficult and often impossible. This is often due to deployment difficulties, lack of available sensors, and cost of maintenance. This talk will explore indirect sensing framework that infer information from physical inertial sensing through a combination of empirical, analytical and heuristic models. The talk then explore this framework through three projects. 1) The SensorFly system, a low-cost, miniature aerial sensor network that aims to be autonomous in deployment, maintenance and adaptation to the environment. By modeling the motion and movement, these sensors collaboratively localize and navigate indoors using inertial sensors. 2) Landing the sensors inside the building, the same sensor sets can be used to discover occupant information including location, identity, status, etc. 3) The same sensors on the human body, the system can inference muscle activity and fatigue level. The talk will explore the commonalities of data understanding and inferencing. Bio: Pei Zhang is an associate research professor in the ECE departments at Carnegie Mellon University. He received his bachelor's degree with honors from California Institute of Technology in 2002, and his Ph.D. degree in Electrical Engineering from Princeton University in 2008. While at Princeton University, he developed the ZebraNet system, which is used to track zebras in Kenya. It was the first deployed, wireless, ad- hoc, mobile sensor network. His recent work includes SensorFly (focus on groups of autonomous miniature-helicopter based sensor nodes) and MARS (Muscle Activity Recognition). Beyond research publications, his work has been featured on popular media including CNN, Science Channel, Discovery Channel, CBS News, CNET, Popular Science, BBC Focus, etc. He is also a cofounder of the startup Vibradotech. In addition, he has won several awards including the NSF CAREER, Edith and Martin B. Stein Solar Energy Innovation Award, and a member of the Department of Defense Computer Science Studies Panel. Margaret Martonosi H. T. Adams '35 Professor of Computer Science. Princeton University ■ mrm(a)princeton.edu ■ http://www.princeton.edu/~mrm

1 0