Select a calendar:
Filter March Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Conferences, Lectures, & Seminars
Events for March
-
CS Colloquium: Bistra Dilkina (Georgia Tech) -Challenges in Computational Sustainability
Wed, Mar 01, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Bistra Dilkina, Georgia Tech
Talk Title: Challenges in Computational Sustainability
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
Computational sustainability is a new interdisciplinary research focused on computational problems that arise in the quest for sustainable development. The goal of sustainable development is to balance environmental, economic, and societal factors to "meet the needs of the present without compromising the ability of future generations to meet their own needs." In this talk, I will provide a sample of computational sustainability problems, from the areas of biodiversity conservation, energy, climate and environment monitoring. I will describe, for example, network design problems motivated by challenging planning problems in wildlife conservation. In this context, I will present a network design optimization framework for stochastic diffusion processes, such as species dispersal, fire spread, information propagation, and disease outbreak. I will also emphasize the unique opportunities for scalable constraint reasoning and optimization techniques to contribute to the new research
area of computational sustainability and describe our recent advances in improving the state-of-the-art in large-scale optimization by leveraging machine learning techniques to inform the design of combinatorial search algorithms.
Biography: Bistra Dilkina is an assistant professor in the College of Computing at the Georgia Institute of Technology and a Fellow at the Brook Byers Institute for Sustainable Systems. She received her PhD from Cornell University in 2012, and was a Post-Doctoral associate at the Institute for Computational Sustainability until 2013. Her research focuses on advancing the state of the art in combinatorial optimization techniques for solving real-world large-scale problems, particularly ones that arise in sustainability areas such as biodiversity conservation planning and urban planning. Her work spans discrete optimization, network design, stochastic optimization, and machine learning. She is also the co-director of the Data Science for Social Good (DSSG) Atlanta summer program, which partners student teams with government and nonprofit organizations to help solve some of their most difficult problems using analytics, modeling, prediction and visualization. Bistra has (co-)authored over 30 publications, and has won several awards, including Best Student Paper runner up at KDD 2016, Best Paper of the INFORMS ENRE Section, Lockheed Inspirational Young Faculty Award, Raytheon Faculty Fellowship, and Georgia Power Professor of Excellence Award.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Adish Singla (ETH Zurich) - Learning With and From People
Mon, Mar 06, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Adish Singla, ETH Zurich
Talk Title: Learning With and From People
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
People are becoming an integral part of computational systems, fueled primarily by recent technological advancements as well as deep-seated economic and societal changes. Consequently, there is a pressing need to design new data science and machine learning frameworks that can tackle challenges arising from human participation (e.g. questions about incentives and users' privacy) and can leverage people's capabilities (e.g. ability to learn).
In this talk, I will share my research efforts at the confluence of people and computing to address real-world problems. Specifically, I will focus on collaborative consumption systems (e.g. shared mobility systems and sharing economy marketplaces like Airbnb) and showcase the need to actively engage users for shaping the demand who would otherwise act primarily in their own interest. The main idea of engaging users is to incentivize them to switch to alternate choices that would improve the system's effectiveness. To offer optimized incentives, I will present novel multi-armed bandit algorithms and online learning methods in structured spaces for learning users' costs for switching between different pairs of available choices. Furthermore, to tackle the challenges of data sparsity and to speed up learning, I will introduce hemimetrics as a structural constraint over users' preferences. I will show experimental results of applying the proposed algorithms on two real-world applications: incentivizing users to explore unreviewed hosts on services like Airbnb and tackling the imbalance problem in bike sharing systems. In collaboration with an ETH Zurich spinoff and a public transport operator in the city of Mainz, Germany, we deployed these algorithms via a smartphone app among users of a bike sharing system. I will share the findings from this deployment.
Biography: Adish Singla is a PhD student in the Learning and Adaptive Systems Group at ETH Zurich. His research focuses on designing new machine learning frameworks and developing algorithmic techniques, particularly for situations where people are an integral part of computational systems. Before starting his PhD, he worked as a Senior Development Lead in Bing Search for over three years. He is a recipient of the Facebook Fellowship in the area of Machine Learning, Microsoft Research Tech Transfer Award, and Microsoft Gold Star Award.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Philip Thomas (CMU) - Safe Machine Learning
Tue, Mar 07, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Philip Thomas, Carnegie Mellon University
Talk Title: Safe Machine Learning
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
Machine learning algorithms are everywhere, ranging from simple data analysis and pattern recognition tools used across the sciences to complex systems that achieve super-human performance on various tasks. Ensuring that they are safe-”that they do not, for example, cause harm to humans or act in a racist or sexist way-”is therefore not a hypothetical problem to be dealt with in the future, but a pressing one that we can and should address now.
In this talk I will discuss some of my recent efforts to develop safe machine learning algorithms, and particularly safe reinforcement learning algorithms, which can be responsibly applied to high-risk applications. I will focus on a specific research problem that is central to the design of safe reinforcement learning algorithms: accurately predicting how well a policy would perform if it were to be used, given data collected from the deployment of a different policy. Solutions to this problem provide a way to determine that a newly proposed policy would be dangerous to use without requiring the dangerous policy to ever actually be used.
Biography: Philip Thomas is a postdoctoral research fellow in the Computer Science Department at Carnegie Mellon University, advised by Emma Brunskill. He received his Ph.D. from the College of Information and Computer Sciences at the University of Massachusetts Amherst in 2015, where he was advised by Andrew Barto. Prior to that, Philip received his B.S. and M.S. in computer science from Case Western Reserve University in 2008 and 2009, respectively, where Michael Branicky was his adviser. Philip's research interests are in machine learning with emphases on reinforcement learning, safety, and designing algorithms that have practical theoretical guarantees.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: David Naylor (CMU) - Privacy in the Internet (Without Giving up Everything Else)
Thu, Mar 09, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: David Naylor, Carnegie Mellon University
Talk Title: Privacy in the Internet (Without Giving up Everything Else)
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
Using the Internet inherently entails privacy risks. Each packet, potentially carrying information that users would rather keep private, is exposed to a network infrastructure operated by a number of third parties the user may not trust and likely cannot even identify. In some cases, the user may not even trust the recipient.
Techniques exist to protect user privacy, but they typically do so at the expense of other desirable properties. For example, anonymity services like Tor hide a packet's true sender, but weaken accountability by making it difficult for network administrators or law enforcement to track down malicious senders. Similarly, encryption hides application data from third parties, but prevents the use of middleboxes---devices that process packets in the network to improve performance (like caches) or security (like intrusion detection systems).
In this talk, I'll present techniques for managing these "Privacy vs. X" conflicts, including a new network architecture that re-thinks basic networking building blocks like packet source addresses and new secure communication protocols that explicitly balance data privacy with the benefits of middleboxes.
Biography: David is a Ph.D. student at Carnegie Mellon University, where he is advised by Peter Steenkiste. His primary research interests are computer networking, security, and privacy, but he is also interested in Web measurement and performance (http://isthewebhttp2yet.com and https://eyeorg.net). David received his B.S. from the University of Iowa in 2011, where he created the DDR inspired "Scrub Scrub Revolution," a handwashing training game for healthcare professionals. He is an NDSEG fellow and received an ACM SIGCOMM best paper award.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Invited Lecture: Richard Anfang - Life Lessons of a Wall St. CIO
Fri, Mar 10, 2017 @ 10:00 AM - 10:50 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Richard Anfang,
Talk Title: Life Lessons of a Wall St. CIO
Abstract: Please join special Guest Lecturer, CIO Richard Anfang, as he shares technology experience and insight from his 30+ year career on Wall Street. Open to all CS students
Richard will discuss technology, innovation, business strategy, and talent. He will also provide valuable career advice and his thoughts on the importance of mentorship.
Biography: Richard Anfang is a technology executive with over 30 years of experience working in global financial service organizations. He has held senior management positions as a business-aligned Chief Information Officer, managed enterprise technology infrastructure, and partnered closely with C-level executives. Anfang has an extensive track record delivering innovative technology solutions to solve business problems within the financial services sector.
Most recently, he worked at JPMorgan Chase where he was Chief Architect of the firm's Global Technology Infrastructure organization. From 2013 to early 2014, he was the Chief Information Officer of JPMorgan's Asset Management business. He was responsible for the business' strategic leadership of technology, overseeing over 3,000 technologists globally spanning the firm's global Investment Management and Private Bank businesses. From 2008 to 2012, he was the Chief Information Officer of JPMorgan's Worldwide Securities Services business.
Prior to joining JPMorgan, he spent over 24 years at Morgan Stanley. During that time, he held a number of senior Information Technology positions including Chief Technology Officer of the Prime Brokerage business, global head of the firm's Enterprise Infrastructure group, head of Equity and Fixed Income Sales & Trading applications development, and Chief Information Officer of Morgan Stanley's UK and European businesses.
Anfang has an extensive track record of delivering innovative solutions and solving business problems in the financial services industry. He has served on the advisory board of several technology service providers and has participated in numerous industry forums and committees.
He holds a Master of Science degree in Computer Science from the University of Pennsylvania (1983) and a Bachelor of Science degree in Computer Science from the University of Michigan (1981).
Host: CS Department
Location: Grace Ford Salvatori Hall Of Letters, Arts & Sciences (GFS) - 116
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Yahoo! Machine Learning Seminar: Anshumali Shrivastava (Rice University) - Probabilistic Hashing for Scalable, Sustainable and Secure Machine Learning
Fri, Mar 17, 2017 @ 10:30 AM - 11:30 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Anshumali Shrivastava, Rice University
Talk Title: Probabilistic Hashing for Scalable, Sustainable and Secure Machine Learning
Series: Yahoo! Labs Machine Learning Seminar Series
Abstract: Large scale machine learning and data mining applications are constantly dealing with datasets at TB scale and the anticipation is that soon it will reach PB level. At this scale, simple data mining operations such as search, learning, and clustering become challenging.
In this talk, we will start with a basic introduction to probabilistic hashing (or fingerprinting) and the classical LSH algorithm. Then I will present some of my recent adventures with probabilistic hashing in making large-scale machine learning practical. I will show how the
idea of probabilistic hashing can be used to significantly reduce the computations in classical machine learning algorithms such Deep Learning (using our recent success with asymmetric hashing for inner products). I will highlight the computational bottleneck, i.e. the hashing time, and will show an efficient variant of minwise hashing. In the end, if time permits, I will demonstrate the use of probabilistic hashing for obtaining practical privacy-preserving
algorithms.
Biography: Anshumali Shrivastava is an assistant professor in the computer science department at Rice University. His broad research interests include large scale machine learning, randomized algorithms for big data systems and graph mining. He is a recipient of 2017 NSF CAREER Award. His research on hashing inner products has won Best Paper Award at NIPS 2014 while his work on representing graphs got the Best Paper Award at IEEE/ACM ASONAM 2014. He obtained his PhD in computer science from Cornell University in 2015.
Host: Yan Liu
Location: Ronald Tutor Hall of Engineering (RTH) - 526
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Yuanjie Li (UCLA) - Stimulating Intelligence in the Mobile Networked Systems
Mon, Mar 20, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Yuanjie Li, UCLA
Talk Title: Stimulating Intelligence in the Mobile Networked Systems
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
The mobile networked systems (4G and upcoming 5G) are at a critical stage of the technology revolution. Despite offering working solutions for billions of users, they are complex and closed: The infrastructure lacks guarantees for the right designs and operations, while the mobile client lacks the insights of the "black-box" network behaviors. Both fundamentally limit our understanding of why various problems could happen, and how to resolve them.
In this talk, I describe primitives that stimulate more infrastructure and client intelligence. For the infrastructure, I present verification and state management techniques that enforce provably correct designs and operations. For the client, I show how a data-driven system design allows it to be more active in improving its performance, reliability, and security. These results suggest that the future systems (5G) should be equipped with more intelligence, and make themselves easy to understand and use.
Biography: Yuanjie Li is a Ph.D. candidate in Computer Science at UCLA, advised by Professor Songwu Lu. His interests include the networked systems, mobile computing, and their security. He has won ACM MobiCom'16 Best Community Paper Award and UCLA Dissertation Year Fellowship in 2016. His work has resulted in an open-source community tool (MobileInsight) used by 130 universities and companies so far.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Justin Cheng (Stanford) - Antisocial Computing: Explaining and Predicting Negative Behavior Online
Tue, Mar 21, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Justin Cheng, Stanford University
Talk Title: Antisocial Computing: Explaining and Predicting Negative Behavior Online
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
Antisocial behavior and misinformation are increasingly prevalent online. As users interact with one another on social platforms, negative interactions can cascade, resulting in complex changes in behavior that are difficult to predict. My research introduces computational methods for explaining the causes of such negative behavior and for predicting its spread in online communities. It complements data mining with crowdsourcing, which enables both large-scale analysis that is ecologically valid and experiments that establish causality. First, in contrast to past literature which has characterized trolling as confined to a vocal, antisocial minority, I instead demonstrate that ordinary individuals, under the right circumstances, can become trolls, and that this behavior can percolate and escalate through a community. Second, despite prior work arguing that such behavioral and informational cascades are fundamentally unpredictable, I demonstrate how their future growth can be reliably predicted. Through revealing the mechanisms of antisocial behavior online, my work explores a future where systems can better mediate interpersonal interactions and instead promote the spread of positive norms in communities.
Biography: Justin Cheng is a PhD candidate in the Computer Science Department at Stanford University, where he is advised by Jure Leskovec and Michael Bernstein. His research lies at the intersection of data science and human-computer interaction, and focuses on cascading behavior in social networks. This work has received a best paper award, as well as several best paper nominations at CHI, CSCW, and ICWSM. He is also a recipient of a Microsoft Research PhD Fellowship and a Stanford Graduate Fellowship.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Nihar Shah (UC Berkeley) - Learning from People
Tue, Mar 21, 2017 @ 04:00 PM - 05:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Nihar Shah, UC Berkeley
Talk Title: Learning from People
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
Learning from people represents a new and expanding frontier for data science. Two critical challenges in this domain are of developing algorithms for robust learning and designing incentive mechanisms for eliciting high-quality data. In this talk, I describe progress on these challenges in the context of two canonical settings, namely those of ranking and classification. In addressing the first challenge, I introduce a class of "permutation-based" models that are considerably richer than classical models, and present algorithms for estimation that are both rate-optimal and significantly more robust than prior state-of-the-art methods. I also discuss how these estimators automatically adapt and are simultaneously also rate-optimal over the classical models, thereby enjoying a surprising a win-win in the bias-variance tradeoff. As for the second challenge, I present a class of "multiplicative" incentive mechanisms, and show that they are the unique mechanisms that can guarantee honest responses. Extensive experiments on a popular crowdsourcing platform reveal that the theoretical guarantees of robustness and efficiency indeed translate to practice, yielding several-fold improvements over prior art.
Biography: Nihar B. Shah is a PhD candidate in the EECS department at the University of California, Berkeley. He is the recipient of the Microsoft Research PhD Fellowship 2014-16, the Berkeley Fellowship 2011-13, the IEEE Data Storage Best Paper and Best Student Paper Awards for the years 2011/2012, and the SVC Aiya Medal from the Indian Institute of Science for the best master's thesis in the department. His research interests include statistics and machine learning, with a current focus on applications to learning from people.
Host: CS Department
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Long Lu (Stony Brook University) - New OS and Programming Support for Securing Mobile and IoT Platforms
Thu, Mar 23, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Long Lu, Stony Brook University
Talk Title: New OS and Programming Support for Securing Mobile and IoT Platforms
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
Software running on mobile and IoT platforms increasingly falls victim to new attacks, which cause device compromises and privacy leaks that are often more severe than their counterparts on conventional computers. My research finds that new attacks on these platforms are possible primarily due to a gap between the evolving security needs of software and the legacy security support provided by operating systems and programming tools.
In this talk, I will first overview my recent works that aim to bridge this gap by rethinking the principles and designs of security mechanisms in operating systems, compilation toolchains, and TEEs (Trusted Execution Environments). I will then present two systems that address a critical yet previously unmet security need of today's apps, namely in-app isolation. The first system introduces a new OS-managed code execution unit, called shred, to compensate thread and process. A shred is a segment of a thread execution. Code inside a shred can access, in addition to the regular virtual memory, a private memory region. Using shreds, programmers can now protect sensitive in-memory code and data against untrusted code running in the same process or thread. The second system enables comprehensive security policy enforcement at the sub-app granularities, preventing mutually distrusting app modules from abusing each other's resources and privileges. In the final part of the talk, I will discuss my ongoing and future works on laying the system foundation for securing IoT platforms.
Biography: Long Lu is an Assistant Professor of Computer Science and the director of RiS3 Lab at Stony Brook University. Long's research spans the broad area of systems and software security. His recent work is focused on application and operating system security for emerging platforms, such as mobile and IoT/CPS devices. He designs code and data protection mechanisms, program analysis techniques, and user-facing software tools to prevent real attacks. He constantly publishes in the top-tier computer security conferences and is frequently invited to serve on their program committees. His research outcomes have been adopted by IBM, Microsoft, NEC, and Samsung. His work is currently funded by NSF, ONR, ARO, and AFRL. Long is a recipient of the NSF CAREER Award and the Air Force Faculty Fellowship. He holds a Ph.D. in Computer Science from Georgia Tech.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Matthew Brown (UCLA) -Typed Self-Applicable Meta-Programming
Thu, Mar 23, 2017 @ 04:00 PM - 05:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Matthew Brown, UCLA
Talk Title: Typed Self-Applicable Meta-Programming
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
Meta-programming is a fundamental technique in computer science. It allows high levels of abstraction to be utilized with low cost. Meta-programs like compilers, interpreters, and program optimizers make high-level programming languages efficient, providing increased programmer productivity and performance comparable to lower-level languages. Self-applicable meta-programming makes meta-programming first-class, enabling many powerful
techniques. However, meta-programming and particularly self-applicable meta-programming is often complex, error-prone and difficult to debug. For these reasons it has untapped potential to provide benefits in many areas. Typed meta-programming uses modern techniques for type checking meta-programs to make them less error-prone and easier to understand and debug. It also brings the power of self-applicable meta-programming to statically-typed languages, ending a long-persisting trade-off between static and dynamic type checking. In this talk I discuss foundational results in typed self-applicable meta-programming.
Biography: Matt Brown is PhD candidate at UCLA, working in the compilers lab under Jens Palsberg. He holds a Bachelor's degree from UC Santa Cruz and a Master's from UCLA. His research focus is typed self-applicable meta-programming, which uses typed program representation techniques to ensure correctness properties of self-applicable meta-programs like self-interpreters. Other research interests include type systems, program verification, concurrency, and functional programming languages. He was recently a part-time lecturer at Loyola Marymount University and has six years of industry experience.
Host: CS Department
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: David Held (UC Berkeley) - Robots in Clutter: Learning to Understand Environmental Changes
Mon, Mar 27, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: David Held, UC Berkeley
Talk Title: Robots in Clutter: Learning to Understand Environmental Changes
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
Robots today are confined to operate in relatively simple, controlled environments. One reason for this is that current methods for processing visual data tend to break down when faced with occlusions, viewpoint changes, poor lighting, and other challenging but common situations that occur when robots are placed in the real world. I will show that we can train robots to handle these variations by modeling the causes behind visual appearance changes. If robots can learn how the world changes over time, they can be robust to the types of changes that objects often undergo. I demonstrate this idea in the context of autonomous driving, and I will show how we can use this idea to improve performance for every step of the robotic perception pipeline: object segmentation, tracking, and velocity estimation. I will also present some recent work on learning to manipulate objects, using a similar framework of learning environmental changes. By learning how the environment can change over time, we can enable robots to operate in the complex, cluttered environments of our daily lives.
Biography: David Held is a post-doctoral researcher at U.C. Berkeley working with Pieter Abbeel on deep reinforcement learning for robotics. He recently completed his Ph.D. in Computer Science at Stanford University with Sebastian Thrun and Silvio Savarese, where he developed methods for perception for autonomous vehicles. David has also worked as an intern on Google's self-driving car team. Before Stanford, David was a researcher at the Weizmann Institute, where he worked on building a robotic octopus. He received a B.S. and M.S. in Mechanical Engineering at MIT and an M.S. in Computer Science at Stanford, for which he was awarded the Best Master's Thesis Award from the Computer Science Department.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Yeongjin Jang (Georgia Tech) - Protecting Computing System Interactions
Tue, Mar 28, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Yeongjin Jang, Georgia Tech
Talk Title: Protecting Computing System Interactions
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
Computing platforms are evolving from desktops to Smartphones to the Internet of things (IoT) devices. In this change, computer systems have started embedding an amazing variety of interaction points in both software and hardware forms. While such changes have made everyday life easier by enabling various convenient features, protecting these systems has become much more difficult. This is not only because system complexity has increased with the integration of more interactions and often conflicts with the existing security mechanisms, but also because improper security practices or incomplete security checks result from faster production cycles that generally lead to more security holes.
In this talk, Yeongjin will present his research on protection of computing system interactions. First, he will present Gyrus, a user interaction monitoring system that reflects user's intention to network traffic monitoring. Gyrus can protect user-to-network interactions such as sending message online and online banking. Next, he will present a result of security analysis on user I/O in operating systems,
in which he discovered computer accessibility as a new attack vector. The analysis found 12 new attacks in popular operating systems, and he discusses countermeasures against the vulnerabilities to keep the affected systems secure.
Biography: Yeongjin Jang is a Ph.D. candidate in Computer Science at the Georgia Institute of Technology. His research focuses on security and privacy problems of computing systems, which include operating systems, mobile systems, and computing hardware.
His research results are recognized for their highly practical impact, as noted by one award and two nominations for the CSAW best applied research paper. Moreover, his research has been widely covered in popular media including Forbes, Wired, MIT Technology Review, and more.
Yeongjin received his M.S. from Georgia Tech in 2016 and B.S. from KAIST in 2010.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Kevin Jamieson (UC Berkeley) - Efficient scalable algorithms for adaptive data collection
Thu, Mar 30, 2017 @ 11:00 AM - 12:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Kevin Jamieson, UC Berkeley
Talk Title: Efficient scalable algorithms for adaptive data collection
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
In many applications, data-driven discovery is limited by the rate of data collection: the skilled labor it takes to operate a pipette, the time to execute a long-running physics simulation, the patience of an infant to remain still in an MRI, or the cost of labeling large corpuses of complex images. A powerful paradigm to extract the most information with such limited resources is active learning, or adaptive data collection, which leverages already-collected data to guide future measurements in a closed loop. But being convinced that data-collection should be adaptive is not the same thing as knowing how to adapt in a way that is both sample efficient and reliable. In this talk, I will present several examples of my provably reliable -- and practical -- adaptive data collection algorithms being applied in the real-world. In particular, I will show how my adaptive algorithms are used each week to crowd-source the winner of the New Yorker Magazine Cartoon Caption Contest. I will also discuss my application of adaptive learning concepts at Google to accelerate the tuning of deep networks in a highly parallelized environment of thousands of GPUs.
Biography: Kevin Jamieson is a postdoctoral researcher working with Professor Benjamin Recht in the Department of Electrical Engineering and Computer Sciences at the University of California, Berkeley. He is interested in the theory and practice of machine learning algorithms that sequentially collect data using an adaptive strategy. This includes active learning, multi-armed bandit problems, and stochastic optimization. Kevin received his Ph.D. from the University of Wisconsin - Madison under the advisement of Robert Nowak. Prior to his doctoral work, Kevin received his B.S. from the University of Washington, and an M.S. from Columbia University, both in electrical engineering.
Host: CS Department
Location: Ronald Tutor Hall of Engineering (RTH) - 217
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Jyotirmoy V. Deshmukh (Toyota Technical Center) -Ninja Temporal Logic: Making formal methods relevant in engineering practice
Thu, Mar 30, 2017 @ 04:00 PM - 05:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Jyotirmoy V. Deshmukh, Toyota Technical Center
Talk Title: Ninja Temporal Logic: Making formal methods relevant in engineering practice
Series: CS Colloquium
Abstract: This lecture satisfies requirements for CSCI 591: Computer Science Research Colloquium.
The software that controls the operation of critical systems such as vehicles, medical devices, buildings, and transportation infrastructures is getting smarter due to the increased demands for autonomy. The push for increased automation is a worthy goal, but can we do so without compromising the safety and reliability of such systems?
Furthermore, can formal methods truly improve a design engineer's productivity? In this talk, I will introduce some of the most important questions facing academic and industrial development of software for the cyber-physical systems of tomorrow. We will consider solutions based on the use of formal logics, that, on one hand allow rigorous reasoning about system designs, while on the other, do not place an undue burden on the engineer. In particular, I will explain how formal requirements using real-time temporal logics have had an impact in the development of cutting-edge alternate-energy vehicles and advanced control problems within Toyota. I will guide the audience through an ecosystem built around temporal logic that permits automatic testing, efficient monitoring, requirement engineering and controller synthesis for highly complex automotive systems. The talk covers topics from what I consider the trifecta for designing reliable cyber-physical systems: formal logic, machine learning, and control theory, and will lay out my vision for future research and open problems within this domain.
Biography: Jyotirmoy V. Deshmukh is a Principal Engineer at Toyota R&D. He received his Ph.D. from the University of Texas at Austin under the supervision of E. Allen Emerson on topics including tree automata, verifying data structure libraries, static analysis for concurrent programs and program repair. He worked as a post-doctoral researcher at the University of Pennsylvania with Rajeev Alur's research group, investigating theoretical models of streaming computation and program synthesis techniques. For the last five years at Toyota, Jyo's research has focused on the design and analysis of industrial cyber-physical systems. Drawing on areas such as hybrid systems, real-time temporal logics, control theory, machine learning and dynamical systems theory, Jyo has been attempting to bridge the gap between academic research and its applicability to industrial-scale systems.
Host: CS Department
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
