- talks - lists.cs.princeton.edu

Guangda Hu will present his FPO, "Lower Bounds for Error-Correcting Codes with Local Recovery" on Monday, 12/12/2016 at 2:00pm in CS 302.
by Nicki Gotsis 05 Dec '16

05 Dec '16

Guangda Hu will present his FPO, "Lower Bounds for Error-Correcting Codes with Local Recovery" on Monday, 12/12/2016 at 2:00pm in CS 302. The members of his committee are: Zeev Dvir (adviser), Readers: Sanjeev Arora and Shubhangi Saraf (Rutgers University); Nonreaders: Mark Braverman and Elad Hazan. A copy of his thesis is available in Room 310. Everyone is invited to attend his talk. The talk abstract follows below. Abstract: Error-correcting codes (ECCs) are ubiquitous in computer science. A common property of ECCs is local recovery, which demands that given a corrupted codeword, a single lost code symbol can be recovered by reading only a small part of the codeword. An intriguing problem is to find the most “efficient” ECCs (e.g., codes with short length, codes over a small alphabet) with certain types of local recovery. Both constructions and lower bounds have been proven in the literature. However, the problem is still largely open. In this thesis, we prove three lower bound results on different types of ECCs with local recovery: Firstly, we propose an approximate version of locally decodable codes (LDCs) and prove lower bounds that are similar to the known ones for traditional LDCs. The concerned approximate LDCs are over real numbers and they support recoveries by querying constant number of codeword symbols. The 2-query case (the bulk of our work) is partially related to the lower bound of constant query LDCs, which is a major open problem. Secondly, we generalize the Sylvester-Gallai (SG) theorem to a subspace version. Generally speaking, the setting of the SG theorem is equivalent to 2-query locally correctable codes (LCCs), and our generalization corresponds to the block version of 2-query LCCs. Thirdly, we consider a realistic storage model that is a unification of several families of codes studied in the literature. We prove negative results for codes that attain the maximal recovering capability under this model. Our lower bound rules out the possibility of constructions of efficient codes for most parameter settings. We will also explore some results in the construction direction in the appendix.

1 0

Soumyadeep Ghosh will present his FPO, "TrustGuard: A Containment Architecture with Verified Output" on Monday, 12/12/2016 at 8am in CS 302.
by Nicki Gotsis 05 Dec '16

05 Dec '16

Soumyadeep Ghosh will present his FPO, "TrustGuard: A Containment Architecture with Verified Output" on Monday, 12/12/2016 at 8am in CS 302. The members of his committee are: David August (adviser), Readers: Andrew Appel and Aarti Gupta; Nonreaders: David Wentzlaff and Simha Sethumadhavan (Columbia University). A copy of his thesis is available in Room 310. Everyone is invited to attend his talk. The talk abstract follows below. Abstract: Computers today are so complex and opaque that a user cannot know everything occurring within the system. Most efforts toward computer security have focused on securing software. However, software security techniques implicitly assume correct execution by the underlying system, including the hardware. Securing these systems has been challenging due to their complexity and the proportionate attack surface they present during their design, manufacturing, deployment, and operation. Ultimately, the user’s trust in the system depends on claims made by each party supplying the system’s components. This dissertation presents the Containment Architecture with Verified Output (CAVO) model in recognition of the reality that existing tools and techniques are insufficient to secure complex hardware components in modern computing systems. Rather than attempt to secure each complex hardware component individually, CAVO establishes trust in hardware using a single, simple, separately manufactured component, called the Sentry. The Sentry bridges a physical gap between the untrusted system and its external interfaces and contains the effects of malicious behavior by untrusted system components before the external manifestation of any such effects. Thus, only the Sentry and the physical gap must be secured in order to assure users of the containment of malicious behavior. The simplicity and pluggability of CAVO’s Sentry enable suppliers and consumers to take additional measures to secure it, including formal verification, supervised manufacture, and supply chain diversification. This dissertation also presents TrustGuard—the first prototype CAVO design—to demonstrate the feasibility of the CAVO model. TrustGuard achieves containment by only allowing the communication of correctly executed results of signed software. The Sentry in TrustGuard leverages execution information obtained from the untrusted processor to enable efficient checking of the untrusted system’s work, even when the Sentry itself is simpler and much slower than the untrusted processor. Simulations show that TrustGuard can guarantee containment of malicious hardware components with a geomean of 8.5% decline in the processor’s performance, even when the Sentry operates at half the clock frequency of the complex, untrusted processor.

1 0

Colloquium Speaker: Gala Yadgar, Technion Institute, Wed, Dec 7, 4:30pm
by Nicole E. Wagenblast 02 Dec '16

02 Dec '16

Colloquium Speaker Dr. Gala Yadgar , from Technion Institute Wednesday, December 7, 2016 - 4:30pm Computer Science 105 The devil is in the details: on the real costs and benefits of WOM codes for Flash NAND flash, used in modern SSDs, is a write-once medium, where each memory cell must be erased prior to writing. The lifetime of an SSD is limited by the number of erasures allowed on each cell. Thus, minimizing erasures is a key objective in SSD design. A promising approach to eliminate erasures and extend SSD lifetime is to use write-once memory (WOM) codes, designed to accommodate additional writes on write-once media. However, these codes inflate the physically stored data by at least 29%, require an extra read operation before each additional write, and their applicability to MLC and TLC flash is restricted. I will describe the first evaluation of the possible benefit from reusing flash pages with WOM codes on real flash chips and an end-to-end FTL design and implementation. Our results expose a considerable gap between the theoretical benefits of page reuse and those that can be achieved on current state-of-the-art hardware. However, they also lay the ground for promising future improvements. Joint work with Fabio Margaglia, Eitan Yaakobi, Yue Li, Assaf Schuster and Andre Brinkmann.

1 0

EE Department Colloquium - Korhammer Lecture Series - December 2, 2016
by Nicole E. Wagenblast 30 Nov '16

30 Nov '16

Korhammer Lecture Series Silicon is the New Steel: Building the Internet of Everything - the World's First Terascale Network Thomas Lee Professor Electrical Engineering, Stanford University December 2 , 2016 5:00 pm - 6:00 pm Bowen Hall Auditorium Reception to follow Steel transformed civilization in the 20th century, shifting from high-tech material to commodity in the process. Silicon is undergoing an analogous transition, as the action shifts from circuits to systems. This talk will argue that multiple convergent trends are pushing us toward the terascale age, presenting us with both historic opportunities and historic challenges. The latter extend from DC to the millimeter wave, and from design tools to the economics of test. Securing a network possessing an "attack surface" of unprecedented magnitude, as well as supplying power to a trillion devices, remain challenges as well. Solving these problems will complete the transition of silicon from today's ubiquity to tomorrow's invisibility. Thomas Lee received his degrees from MIT. His 1989 doctoral thesis described the world's first CMOS radio. He has been at Stanford University since 1994, having previously worked at Analog Devices, Rambus and other companies. He's helped design PLLs for several microprocessors (notably AMD's K6-K7-K8 and DEC's StrongARM), and has founded or cofounded several companies, including 3D memory company, Matrix Semiconductor (acquired by Sandisk), and IoE companies ZeroG Wireless (acquired by Microchip) and Ayla Networks. He is an IEEE and Packard Foundation Fellow, has won "Best Paper" awards at CICC and ISSCC, and was awarded the 2011 Ho-Am Prize in Engineering. He is a past Director of DARPA's Microsystems Technology Office, serves on the board of Xilinx, and owns between 100 and 200 oscilloscopes, thousands of vacuum tubes, and kilograms of obsolete semiconductors. No one, including himself, quite knows why. 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 Save

1 0

TODAY 4:30pm: CITP/PIIRS GSR public lecture: "Internet Resilience: Global Network, Global Risks" with Jennifer Rexford & Nick Feamster, Wed, Nov 30, 4:30pm, CS 105
by Nicole E. Wagenblast 30 Nov '16

30 Nov '16

This afternoon at 4:30pm in Computer Science 105 (Small Auditorium), the Center for Information Technology Policy (CITP) and the PIIRS Global Systemic Risk research community are co-hosting a public lecture and discussion on "Internet Resilience: Global Network, Global Risks" with Jennifer Rexford and Nick Feamster. Jen is Chair of the Computer Science Department at Princeton and Nick is Professor of Computer Science and Acting Director of CITP. In the study of systemic risk, the Internet is an informative case study as it is our most recently created system, consciously designed and constructed with the goal of avoiding fragility and failure. Internet resilience has emerged as an increasingly critical area of study, as more and more of our systems—finance, infrastructure, transportation, health care, agriculture, energy & electricity, among others—are interconnected with and interdependent upon the Internet. In addition to the impact of the Internet on these macro systems that support our lives, we are increasingly affected by Internet security on an individual level, as many of our consumer products—such as smartphones, refrigerators, thermostats, and cars—are becoming connected to and reliant upon the Internet in the emerging "Internet of Things" (IoT). Jen and Nick will discuss architecture and design of the Internet, how it operates, security risks, and instances of attacks and failures. One topic of discussion will be security threats posed by IoT addressed in recent CITP research ( news and report ). Brief bios for Jen and Nick and an abstract for their talk are below. The lecture will be held this afternoon, Wednesday, November 30th, at 4:30pm in Computer Science 105 (Small Auditorium), and will be followed by Q&A and discussion. The talk is open to the public—please share this invitation with members of your department and all those you think may be interested. ***************** Abstract: The Internet has rapidly become a critical part of the world's communications infrastructure. Yet this global “network of networks” is remarkably vulnerable to cyberattacks, physical damage, and manipulation of the user experience through censorship, wiretapping, and propaganda. In this public discussion, we discuss how the Internet consists of tens of thousands of networks owned by different institutions in different countries, with different economic and political goals. We discuss how early decisions in the design of the Internet led to both tremendous innovation and the substantial security and privacy risks. The architecture of the Internet is a case study in the design and construction of resilient networks and systems. However, as more of our global systems—and our lives and livelihoods—become increasingly interdependent on the Internet, the potential fragility of the Internet is an increasing cause for concern and academic study. Jennifer Rexford is the Gordon Y.S. Wu Professor of Engineering, and Chair of the Computer Science Department. Before joining Princeton in 2005, she worked for eight years at AT&T Labs—Research. She is a member of the American Academy of Arts and Sciences and the National Academy of Engineering and was the 2004 winner of ACM's Grace Murray Hopper Award for outstanding young computer professional. She is co-author of the book Web Protocols and Practice (2001). She received her B.S.E. in electrical engineering from Princeton, and her Ph.D. in electrical engineering and computer science from the University of Michigan. Nick Feamster is Professor of Computer Science and Acting Director of the Center for Information Technology Policy (CITP) at Princeton. His research focuses on networked computer systems, with a strong emphasis on network operations, network architecture and protocol design, high performance wired and wireless networks, and anti-censorship techniques and systems. The collective goal of these research areas is to help network operators to manage their networks better and enable users of these networks to experience high availability and end-to-end performance. He received both his B.S. in electrical engineering and computer science and his Ph.D. in computer science from MIT. *************

1 0

Colloquium Speaker: Alexis Battle, Tuesday, today, 12:30pm
by Nicole E. Wagenblast 29 Nov '16

29 Nov '16

Colloquium Speaker Prof. Alexis Battle, Johns Hopkins University Tuesday, November 29, 2016 - 12:30pm Computer Science 105 Predicting the impact of rare regulatory variation The enormous increase in availability of full human genomic sequences presents great opportunity for understanding the impact of rare genetic variants. Based on current knowledge, however, we are still limited in our ability to interpret or predict regulatory consequences of rare variants. Association methods are inherently limited for variants observed in only one or a few individuals. Diverse genomic annotations such as growing resources of epigenetic data and have been shown to be informative regarding regulatory elements, but are only moderately predictive of impact for individual variants. The availability of personal RNA-seq and other cellular measurements for the same individuals with genome sequencing offers a new avenue for integrated methods for prioritizing rare functional variants. By considering informative genomic annotations along with molecular phenotyping, we are able to identify the regulatory impact of rare genetic variants largely excluded from previous analyses. We have evaluated the impact of rare regulatory variants using whole genome sequences and corresponding RNA-sequence in 44 different tissues samples from the GTEx project and report. We demonstrate that rare variants in promoter elements, conserved regions, and variants that affect splicing are associated with extreme changes in gene expression across multiple tissues. Additionally, we have developed a Bayesian machine learning approach that integrates whole genome sequencing with RNA-seq data from the same individual, leveraging gene expression levels along with diverse genomic annotations and performing joint inference to identify likely functional rare regulatory variants. We have applied this model to the GTEx data to prioritize the most likely functional rare regulatory variants for each individual. We demonstrate that integrative models perform better than predictions from DNA-sequencing or RNA-sequencing alone. Our probabilistic model of rare regulatory genetic variants offers potential for identifying deleterious regulatory variants from individual genomes. Alexis Battle’s research focuses on unraveling the impact of genetics on the human health, using machine learning and probabilistic methods to analyze large scale genomic data. She is interested in applications to personal genomics, genetics of gene expression, and gene networks in disease, leveraging diverse data to infer more comprehensive models of genetic effects on the cell. She earned her Ph.D. and Masters in Computer Science in 2014 from Stanford University in 2014, where she also received her Bachelors in Symbolic Systems (2003). Alexis also spent several years in industry as a member of the technical staff at Google. Prior to joining Hopkins, Alexis spent a year as a postdoc with Jonathan Pritchard with HHMI and the Genetics Department at Stanford. She joined JHU in July 2014.

1 0

Katie Anna Wolf will present her Pre-FPO "Techniques and Tools for End-User Sonification Design and Audio Personalization" on Monday, December 5th, 2016 at 10am CS 302.
by Nicki Gotsis 28 Nov '16

28 Nov '16

Katie Anna Wolf will present her Pre-FPO on Monday, December 5th, 2016 at 10am CS 302. The members of her committee are as follows: Examiners: Rebecca Fiebrink (advisor); Dan Truman (Music), and Marshini Chetty. Readers: Adam Finkelstein and Mike Nees (Lafayette College) Everyone is invited to attend her talk. The talk title and abstract follow below: Title: Techniques and Tools for End-User Sonification Design and Audio Personalization Abstract: Modern daily life involves consuming increasing amounts of data from personalized applications and devices such as Twitter, the stock market and health trackers. As the amount and types of data increase, we need to continue to develop technology that presents the information in such a way that its meaning can be interpreted not only by scientists but by everyday people. Sonification, as a sister to visualization, is a way of representing data with sound. Due to physical impairments, high-dimensional data that is hard to visualize, or just a general desire to avoid staring at a screen, audio interfaces have been developed to take advantage of the communicative power of sound. However designing sonifications requires understanding of how the end users will use the system as well as knowledge in many domains including sound design, sonification design, and programming. In this talk I will present methods that give users who are non-experts in sound and sonification design the ability to customize their own sonifications using a graphical user interface. These methods include: (1) the analysis and modeling of environmental soundscapes (including animal vocalizations and sounds of the weather) which allow non-experts to talk about sounds in terms they know, (2) using related work by sonification experts and my own exploration of environmental sounds to design an automatic mapping technique that determines how the data should be represented by sounds, and (3) using an example-based approach that allows users to select example sound sets as an exemplar for how the sonification should sound. I will also discuss my ongoing work in developing tools that allow end users to control and personalize their auditory environment. Finally I will present a survey about people’s experience with audio personalization, and how the results may influence the design of tools that put the end user at the center of the sonification design process.

1 0

Colloquium Speaker: Alexis Battle, Tuesday, Nov 29, 12:30pm- reminder
by Nicole E. Wagenblast 28 Nov '16

28 Nov '16

Colloquium Speaker Prof. Alexis Battle, Johns Hopkins University Tuesday, November 29, 2016 - 12:30pm Computer Science 105 Predicting the impact of rare regulatory variation The enormous increase in availability of full human genomic sequences presents great opportunity for understanding the impact of rare genetic variants. Based on current knowledge, however, we are still limited in our ability to interpret or predict regulatory consequences of rare variants. Association methods are inherently limited for variants observed in only one or a few individuals. Diverse genomic annotations such as growing resources of epigenetic data and have been shown to be informative regarding regulatory elements, but are only moderately predictive of impact for individual variants. The availability of personal RNA-seq and other cellular measurements for the same individuals with genome sequencing offers a new avenue for integrated methods for prioritizing rare functional variants. By considering informative genomic annotations along with molecular phenotyping, we are able to identify the regulatory impact of rare genetic variants largely excluded from previous analyses. We have evaluated the impact of rare regulatory variants using whole genome sequences and corresponding RNA-sequence in 44 different tissues samples from the GTEx project and report. We demonstrate that rare variants in promoter elements, conserved regions, and variants that affect splicing are associated with extreme changes in gene expression across multiple tissues. Additionally, we have developed a Bayesian machine learning approach that integrates whole genome sequencing with RNA-seq data from the same individual, leveraging gene expression levels along with diverse genomic annotations and performing joint inference to identify likely functional rare regulatory variants. We have applied this model to the GTEx data to prioritize the most likely functional rare regulatory variants for each individual. We demonstrate that integrative models perform better than predictions from DNA-sequencing or RNA-sequencing alone. Our probabilistic model of rare regulatory genetic variants offers potential for identifying deleterious regulatory variants from individual genomes. Alexis Battle’s research focuses on unraveling the impact of genetics on the human health, using machine learning and probabilistic methods to analyze large scale genomic data. She is interested in applications to personal genomics, genetics of gene expression, and gene networks in disease, leveraging diverse data to infer more comprehensive models of genetic effects on the cell. She earned her Ph.D. and Masters in Computer Science in 2014 from Stanford University in 2014, where she also received her Bachelors in Symbolic Systems (2003). Alexis also spent several years in industry as a member of the technical staff at Google. Prior to joining Hopkins, Alexis spent a year as a postdoc with Jonathan Pritchard with HHMI and the Genetics Department at Stanford. She joined JHU in July 2014.

1 0

Fwd: Princeton Optimization Seminar, Bernard Chazelle, Thursday, Dec. 1, 4:30 PM, Sherrerd 101
by Nicole E. Wagenblast 28 Nov '16

28 Nov '16

----- Princeton Optimization Seminar ----- DATE: TODAY Thursday, December 1, 2016 TIME : 4:30PM LOCATION: Sherrerd Hall 101 SPEAKER: Bernard Chazelle , Princeton University TITLE: The Mathematics of Natural Algorithms Abstract: I will review some of the recent techniques we've used in our study of natural algorithms. These include Dirichlet series for matrix products, mean-field approximations in opinion dynamics, graph sequence grammars, anti-mixing methods for iterated learning, and tools for renormalizing network-based dynamical systems. The talk will be self-contained and non-technical. Bio: Bernard Chazelle is Eugene Higgins Professor of Computer Science at Princeton University, where he has been on the faculty since 1986. His current research focuses on the “algorithmic nature” of living systems. A professor at the Collège de France in Paris in recent years as well as a member of the Institute for Advanced Study in Princeton, he received his PhD in computer science from Yale University in 1980. The author of the book, "The Discrepancy Method," he is a fellow of the American Academy of Arts and Sciences, the European Academy of Sciences, the Association for Computing Machinery, and the recipients of three Best-Paper awards from SIAM. *** If you would like to subscribe to the mailing list of the optimization seminar series, please visit the link below or send an email with SUBSCRIBE opt-seminar in the body to listserv(a)lists.princeton.edu . https://lists.princeton.edu/cgi-bin/wa?SUBED1=opt-seminar&A=1

1 0

Colloquium Speaker: Alexis Battle, Tuesday, Nov 29, 12:30pm
by Nicole E. Wagenblast 22 Nov '16

22 Nov '16

Colloquium Speaker Prof. Alexis Battle, Johns Hopkins University Tuesday, November 29, 2016 - 12:30pm Computer Science 105 Predicting the impact of rare regulatory variation The enormous increase in availability of full human genomic sequences presents great opportunity for understanding the impact of rare genetic variants. Based on current knowledge, however, we are still limited in our ability to interpret or predict regulatory consequences of rare variants. Association methods are inherently limited for variants observed in only one or a few individuals. Diverse genomic annotations such as growing resources of epigenetic data and have been shown to be informative regarding regulatory elements, but are only moderately predictive of impact for individual variants. The availability of personal RNA-seq and other cellular measurements for the same individuals with genome sequencing offers a new avenue for integrated methods for prioritizing rare functional variants. By considering informative genomic annotations along with molecular phenotyping, we are able to identify the regulatory impact of rare genetic variants largely excluded from previous analyses. We have evaluated the impact of rare regulatory variants using whole genome sequences and corresponding RNA-sequence in 44 different tissues samples from the GTEx project and report. We demonstrate that rare variants in promoter elements, conserved regions, and variants that affect splicing are associated with extreme changes in gene expression across multiple tissues. Additionally, we have developed a Bayesian machine learning approach that integrates whole genome sequencing with RNA-seq data from the same individual, leveraging gene expression levels along with diverse genomic annotations and performing joint inference to identify likely functional rare regulatory variants. We have applied this model to the GTEx data to prioritize the most likely functional rare regulatory variants for each individual. We demonstrate that integrative models perform better than predictions from DNA-sequencing or RNA-sequencing alone. Our probabilistic model of rare regulatory genetic variants offers potential for identifying deleterious regulatory variants from individual genomes. Alexis Battle’s research focuses on unraveling the impact of genetics on the human health, using machine learning and probabilistic methods to analyze large scale genomic data. She is interested in applications to personal genomics, genetics of gene expression, and gene networks in disease, leveraging diverse data to infer more comprehensive models of genetic effects on the cell. She earned her Ph.D. and Masters in Computer Science in 2014 from Stanford University in 2014, where she also received her Bachelors in Symbolic Systems (2003). Alexis also spent several years in industry as a member of the technical staff at Google. Prior to joining Hopkins, Alexis spent a year as a postdoc with Jonathan Pritchard with HHMI and the Genetics Department at Stanford. She joined JHU in July 2014.

1 0