Select a calendar:
Filter March Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Events for March
-
CS Colloquium: Melisa Orta Martinez (Stanford University) - Design and Analysis of Open-Source Educational Haptic Devices
Mon, Mar 02, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Melisa Orta Martinez, Stanford University
Talk Title: Design and Analysis of Open-Source Educational Haptic Devices
Series: CS Colloquium
Abstract: The sense of touch (haptics) is an active perceptual system used from our earliest days to discover the world around us. However, formal education is not designed to take advantage of this sensory modality. As a result, very little is known about the effects of using haptics in K-12 and higher education or the requirements for haptic devices for educational applications. This talk will present three novel, open-source, low-cost haptic devices for educational applications and discuss some general principles for designing such devices. The first device, Hapkit is a one-degree-of-freedom kinesthetic device that has been used in several education environments, where we have discovered the potential of haptics to display abstract mathematical concepts and observed the importance of device customization for the students. The second, Haplink, introduces a novel mechanism that enables the device to transform between a one- and two-degree-of-freedom haptic device in order to enable additive learning. The third device, HapCaps is a tactile haptic device that was developed to study the connection between finger perception and math learning in young children. The aim is to design haptic devices that can be used in several educational environments in order to understand the role of haptics in learning.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Melisa Orta Martinez received the BS degree in electronic systems engineering from the Instituto Tecnologico y de Estudios Superiores de Monterrey in Mexico City, during which she spent a year working as a research intern at the Heinz Nixdorf Institute, Paderborn, Germany. She then obtained a MS degree in electrical engineering from Stanford University, Stanford, CA. After her masters degree she worked at Apple Inc. for three years in the Human Interface Devices group. She is currently working toward the doctoral degree in mechanical engineering at Stanford University. Her research interests include haptics, robotics and education.
Host: Heather Culbertson
Location: Ronald Tutor Hall of Engineering (RTH) - 109
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
PhD Thesis Proposal - Emily Sheng
Tue, Mar 03, 2020 @ 10:00 AM - 11:30 AM
Thomas Lord Department of Computer Science
University Calendar
Title: Towards Fairness in Natural Language Processing
Date/Time: Tuesday, March 3rd, 10-11:30am
Location: SAL 213
Candidate: Emily Sheng
Committee: Prof. Prem Natarajan (advisor), Prof. Nanyun Peng, Prof. Aram Galstyan, Prof. Shri Narayanan, Prof. Yan Liu
Abstract: With the advent of more effective, large-scale natural language processing (NLP) techniques, issues of fairness and bias in NLP techniques have become increasingly important. Biased models have the potential to perpetuate and amplify societal biases, which has implications for ethics, model robustness, and model interpretability. First, we describe our work to define biases in a language generation setting. We subsequently describe how different definitions of bias can be used to analyze biases in existing NLP systems, e.g., language generation and named entity recognition. Finally, we propose techniques that allow us to move towards the mitigation and control of biases. This talk will examine the importance of defining tasks and metrics for biases in NLP, how our bias analyses can inform our approach to bias mitigation, and related directions in how we can move towards fairness in NLP.
Location: Henry Salvatori Computer Science Center (SAL) - 213
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
-
CS Colloquium: Kaiyu Hang (Yale University) - Robotic Manipulation – From Representations to Actions
Tue, Mar 03, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Kaiyu Hang, Yale University
Talk Title: Robotic Manipulation -“ From Representations to Actions
Series: CS Colloquium
Abstract: Dexterous manipulation is a challenging and integral task involving a number of subproblems to be addressed, such as perception, planning, and control. Problem representation, which is an essential element in a system that defines what is actually the problem to be considered, determines both the capability of a system and the feasibility of applying such a system in real tasks.
In this talk, I will introduce how good representations can convert difficult problems into easier ones. In particular, I first discuss the development of representations for grasp optimization, as well as how a good representation can simplify and unify the whole grasping system, including globally optimal grasp planning, sensing, adaptation, and control. By expanding or varying this representation in terms of problem scenarios, I further show how it can greatly facilitate solving other problems, such as grasp-aware motion planning, optimal placement planning, and even dual-arm manipulation. Second, I will introduce our work on underactuated manipulation using soft robotic hands. For underactuated hands without any joint encoders or tactile sensors, I present our representations that can enable a robot to interact with tabletop objects using nonprehensile manipulation to finally grasp it, and show how to register the object into its own hand-object system once grasped, so as to eventually provide precise and dexterous in-hand manipulation. Finally, I discuss how to develop representations for optimizing robot fingertip designs, especially for simple grippers with limited grasping skills. By installing those optimized fingertip designs onto drones, I further show that those optimized designs can enable the drones to perch or rest at various structures, achieving significant improvement in energy consumption.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Kaiyu Hang is a postdoctoral associate working with Prof. Aaron M. Dollar at the GRAB lab, Yale University. He received his Ph.D. in Computer Science, specialized in Robotics and Computer Vision, under the supervision of Prof. Danica Kragic from KTH Royal Institute of Technology, Stockholm, Sweden. Before joining the GRAB lab, he was a research assistant professor at the Department of Computer Science and Engineering, and a Junior Fellow of the Institute for Advanced Study, Hong Kong University of Science and Technology. His research interests include representations and optimization for robotic manipulation, motion planning, adaptive grasping and in-hand manipulation, underactuated robotic hands, dual arm manipulation, and mobile manipulation.
Host: Joseph Lim
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Distinguished Lecture: Oren Etzioni (Allen Institute for AI) - Artificial Intelligence and the Future of Humanity
Tue, Mar 03, 2020 @ 04:00 PM - 05:20 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Oren Etzioni, Allen Institute for AI
Talk Title: Artificial Intelligence and the Future of Humanity
Series: Computer Science Distinguished Lecture Series
Abstract: Could we wake up one morning to find that AI is poised to take over the world? Is AI the technology of unfairness and bias?
My talk will assess these concerns, and sketch a more optimistic view.
We will have ample warning before the emergence of superintelligence, and in the meantime we have the opportunity to create
Beneficial AI:
• AI that mitigates bias rather than amplifying it
• AI that saves lives rather than taking them
• AI that helps us to solve humanity's thorniest problems
My talk builds on work at the Allen Institute for AI, a non-profit research institute based in Seattle.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Oren Etzioni launched the Allen Institute for AI, and has served as its CEO since 2014.
He has been a Professor at the University of Washington's Computer Science department since 1991, publishing papers that have garnered over 2,300 highly influential citations on Semantic Scholar.
He is also the founder of several startups including Farecast (acquired by Microsoft in 2008).
Host: Craig Knoblock and Xiang Ren
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Computer Science Department
-
CS Colloquium: Peng Qi (Stanford University) - Explainable and Efficient Knowledge Acquisition from Text
Wed, Mar 04, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Peng Qi, Stanford University
Talk Title: Explainable and Efficient Knowledge Acquisition from Text
Series: CS Colloquium
Abstract: Human languages have served as the media for our knowledge over generations. With the rise of the digital world, making use of the knowledge that is encoded in text has become unprecedentedly important yet challenging. In recent years, the NLP community has made great progress towards operationalizing textual knowledge by building accurate systems that answer factoid questions. However, largely relying on matching local text patterns, these systems fall short at their ability to perform complex reasoning, which limits our effective use of textual knowledge. To address this problem, I will first talk about two distinct approaches to enable NLP systems to perform multi-step reasoning that is explainable to humans, through extracting facts from natural language and answering multi-step questions directly from text. I will then demonstrate that beyond static question answering with factoids, true informativeness of answers stems from communication. To this end, I will show how we lay the foundation for reasoning about latent information needs in conversations to effectively exchange information beyond providing factoid answers.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Peng Qi is a Computer Science PhD student at Stanford University. His research interests revolve around building natural language processing systems that better bridge between humans and the large amount of textual information we are engulfed in. He is excited about building scalable and explainable AI systems, and has worked on extracting knowledge representations from text, question answering involving complex reasoning, and multi-lingual NLP.
Host: Xiang Ren
Location: Ronald Tutor Hall of Engineering (RTH) - 109
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Emma Pierson (Stanford) - Data Science Methods to Reduce Inequality and Improve Healthcare
Thu, Mar 05, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Emma Pierson, Stanford University
Talk Title: Data Science Methods to Reduce Inequality and Improve Healthcare
Series: CS Colloquium
Abstract: I will describe how to use data science methods to understand and reduce inequality in two domains: criminal justice and healthcare. First, I will discuss how to use Bayesian modeling to detect racial discrimination in policing. Second, I will describe how to use machine learning to explain racial and socioeconomic inequality in pain.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Emma Pierson is a PhD student in Computer Science at Stanford, supported by Hertz and NDSEG Fellowships. Previously, she completed a master's degree in statistics at Oxford on a Rhodes Scholarship. She develops statistical and machine learning methods to study two deeply entwined problems: reducing inequality and improving healthcare. She also writes about these topics for broader audiences in publications including The New York Times, The Washington Post, FiveThirtyEight, and Wired. Her work has been recognized by best paper (AISTATS 2018), best poster (ICML Workshop on Computational Biology), and best talk (ISMB High Throughput Sequencing Workshop) awards, and she has been named a Rising Star in EECS and Forbes 30 Under 30 in Science.
Host: Bistra Dilkina
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Lili Su (MIT) - Learning with Distributed Systems: Adversary-Resilience and Neural Networks
Mon, Mar 09, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Lili Su, MIT
Talk Title: Learning with Distributed Systems: Adversary-Resilience and Neural Networks
Series: CS Colloquium
Abstract: In this talk, I will first talk about how to secure Federated Learning (FL) against adversarial faults.
FL is a new distributed learning paradigm proposed by Google. The goal of FL is to enable the cloud (i.e., the learner) to train a model without collecting the training data from users' mobile devices. Compared with traditional learning, FL suffers serious security issues and several practical constraints call for new security strategies. Towards quantitative and systematic insights into the impacts of those security issues, we formulated and studied the problem of Byzantine-resilient Federated Learning. We proposed two robust learning rules that secure gradient descent against Byzantine faults. The estimation error achieved under our more recently proposed rule is order-optimal in the minimax sense.
Then, I will briefly talk about our recent results on neural networks, including both biological and artificial neural networks. Notably, our results on the artificial neural networks (i.e., training over-parameterized 2-layer neural networks) improved the state-of-the-art. In particular, we showed that nearly-linear network over-parameterization is sufficient for the global convergence of gradient descent.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Lili Su is a postdoc in the Computer Science and Artificial Intelligence Laboratory (CSAIL) at MIT, hosted by Professor Nancy Lynch. She received a Ph.D. in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign in 2017, supervised by Professor Nitin H. Vaidya. Her research intersects distributed systems, learning, security, and brain computing. She was the runner-up for the Best Student Paper Award at DISC 2016, and she received the 2015 Best Student Paper Award at SSS 2015. She received UIUC's Sundaram Seshu International Student Fellowship for 2016, and was invited to participate in Rising Stars in EECS (2018). She has served on TPC for several conferences including ICDCS and ICDCN.
Host: Leana Golubchik
Location: Ronald Tutor Hall of Engineering (RTH) - 109
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
PhD Defense - Johnathan Mell
Mon, Mar 09, 2020 @ 11:00 AM - 01:00 PM
Thomas Lord Department of Computer Science
University Calendar
PhD Candidate: Johnathan Mell
Date: Monday, March 9th, 2020
Time: 11:00 AM - 1:00 PM
Location: SAL 213
Committee: Dr. Jonathan Gratch (Chair), Dr. Nate Fast, Dr. Sven Koenig
Title: A Framework for Research in Human-Agent Negotiation
Abstract:
Increasingly, automated agents are interacting with humans in highly social interactions. Many of these interactions can be characterized as negotiation tasks. There has been broad research in negotiation techniques between humans (in business literatures, e.g.), as well a great deal of work in creating optimal agents that negotiate with each other. However, the creation of effective socially-aware agents requires fundamental basic research on human-agent negotiation. Furthermore, this line of enquiry requires highly customizable, fully-interactive systems that are capable of enabling and implementing human-agent interaction. Previous attempts that rely on hypothetical situations or one-shot studies are insufficient in capturing truly social behavior.
This dissertation showcases my invention and development of the Interactive Arbitration Guide Online (IAGO) platform, which enables rigorous human-agent research. IAGO has been designed from the ground up to embody core principles gleaned from the rich body of research on how people actually negotiate. I demonstrate several examples of how IAGO has already yielded fundamental contributions towards our understanding of human-agent negotiation. I also demonstrate how IAGO has contributed to a community of practice by allowing researchers across the world to easily develop and investigate novel algorithms. Finally, I discuss future plans to use this framework to explore how humans and machines can establish enduring and profitable relationships through repeated negotiations.
Location: Henry Salvatori Computer Science Center (SAL) - 213
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
-
CS Colloquium: Antoine Bosselut (University of Washington) - Neuro-symbolic Representations for Commonsense Knowledge and Reasoning
Tue, Mar 10, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Antoine Bosselut, University of Washington
Talk Title: Neuro-symbolic Representations for Commonsense Knowledge and Reasoning
Series: CS Distinguished Lectures
Abstract: Situations described using natural language are richer than what humans explicitly communicate. For example, the sentence "She pumped her fist" connotes many potential auspicious causes. For machines to understand natural language, they must be able to reason about the commonsense inferences that underlie explicitly stated information. In this talk, I will present work on combining traditional symbolic knowledge and reasoning techniques with modern neural representations to endow machines with these capacities.
First, I will describe COMET, an approach for learning commonsense knowledge about unlimited situations and concepts using transfer learning from language to knowledge. Second, I will demonstrate how these neural knowledge representations can dynamically construct symbolic graphs of contextual commonsense knowledge, and how these graphs can be used for interpretable, generalized reasoning. Finally, I will discuss current and future research directions on conceptualizing NLP as commonsense simulation, and the impact of this framing on challenging open-ended tasks such as story generation.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Antoine Bosselut is a PhD Student at the University of Washington advised by Professor Yejin Choi, and a student researcher at the Allen Institute for Artificial Intelligence. His research focuses on building systems for commonsense knowledge representation and reasoning that combine the strengths of modern neural and traditional symbolic methods. He was also a student researcher on the Deep Learning team at Microsoft Research from 2017 to 2018. He is supported by an AI2 Key Scientific Challenges award.
Host: Xiang Ren
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Jesse Thomason (University of Washington) - Language Grounding with Robots
Wed, Mar 11, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Jesse Thomason, University of Washington
Talk Title: Language Grounding with Robots
Series: CS Colloquium
Abstract: We use language to refer to objects like "toast", "plate", and "table" and to communicate requests such as "Could you make breakfast?" In this talk, I will present work on computational methods to tie language to physical, grounded meaning. Robots are an ideal platform for such work because they can perceive and interact with the world. I will discuss dialog and learning strategies I have developed to enable robots to learn from their human partners, similar to how people learn from one another through interaction. I will present methods enabling robots to understand language referring expressions like "the heavy, metallic mug", the first work showing that it is possible to learn to connect words to their perceptual properties in the visual, tactile, and auditory senses of a physical robot. I will also present benchmarks and models for translating high-level human language like "put the toast on the table" that imply latent, intermediate goals into executable sequences of agent actions with the help of low-level, step-by-step language instructions. Finally, I will discuss how my work in grounded language contributes to NLP, robotics, and the broader goals of the AI community.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Jesse Thomason is a postdoctoral researcher at the University of Washington working with Luke Zettlemoyer. He received his PhD from the University of Texas at Austin with Raymond Mooney. His research focuses on language grounding and natural language processing applications for robotics (RoboNLP). Key to this work is using dialog with humans to facilitate both robot task execution and learning to enable lifelong improvement of robots' language understanding capabilities. He has worked to encourage and promote work in RoboNLP through workshop organization at both NLP and robotics conference venues.
Host: Stefanos Nikolaidis
Location: Ronald Tutor Hall of Engineering (RTH) - 109
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
Computer Science General Faculty Meeting
Wed, Mar 11, 2020 @ 12:00 PM - 02:00 PM
Thomas Lord Department of Computer Science
Receptions & Special Events
Bi-Weekly regular faculty meeting for invited full-time Computer Science faculty only. Event details emailed directly to attendees.
Location: Ronald Tutor Hall of Engineering (RTH) - 526
Audiences: Invited Faculty Only
Contact: Assistant to CS chair
-
*CANCELLED* CAIS Seminar: Rediet Abebe (Harvard University) - Mechanism Design for Social Good
Wed, Mar 11, 2020 @ 04:15 PM - 05:15 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Rediet Abebe, Harvard University
Talk Title: Mechanism Design for Social Good
Series: USC Center for Artificial Intelligence in Society (CAIS) Seminar Series
Abstract: Algorithmic and artificial intelligence techniques show immense potential to deepen our understanding of socioeconomic inequality and inform interventions designed to improve access to opportunity. Interventions aimed at historically under-served communities are made particularly challenging by the fact that disadvantage and inequality are multifaceted, notoriously difficult to measure, and reinforced by feedback loops in underlying structures.
In this talk, we develop algorithmic and computational techniques to address these issues through two types of interventions: one in the form of allocating scarce societal resources and another in the form of improving access to information. We examine the ways in which techniques from algorithms, discrete optimization, and network and computational science can combat different forms of disadvantage, including susceptibility to income shocks, social segregation, and disparities in access to health information. We discuss current practice and policy informed by this work and close with a discussion of an emerging research area -- Mechanism Design for Social Good (MD4SG) -- around the use of algorithms, optimization, and mechanism design to address this category of problems.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Rediet Abebe is a Junior Fellow at the Harvard Society of Fellows and will be receiving her Ph.D. in computer science from Cornell University in 2019. Her research is broadly in the fields of algorithms and AI, with a focus on equity and social good concerns. As part of this research agenda, she co-founded Mechanism Design for Social Good (MD4SG), a multi-institutional, interdisciplinary research initiative working to improve access to opportunity for historically disadvantaged communities. This initiative has active participants from over 100 institutions in 20 countries and has been supported by Schmidt Futures, the MacArthur Foundation, and the Institute for New Economic Thinking.
Abebe currently serves on the NIH Advisory Committee to the Director Working Group on AI, tasked with developing a comprehensive report to the NIH leadership. She was recently named one of 35 Innovators Under 35 by the MIT Technology Review and honored in the 2019 Bloomberg 50 list as a "one to watch." Her work has been covered by outlets including Forbes, the Boston Globe, and the Washington Post. In addition to her research, she also co-founded Black in AI, a non-profit organization tackling diversity and inclusion issues in AI. Her research is deeply influenced by her upbringing in her hometown of Addis Ababa, Ethiopia.
Host: USC Center for Artificial Intelligence in Society (CAIS)
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Computer Science Department
-
*CANCELLED* CAIS Seminar: Meredith Gore - Wildlife Trafficking in the Anthropocene: Conservation, Crime & Communities
Thu, Mar 12, 2020 @ 09:45 AM - 10:45 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Meredith Gore, PhD
Talk Title: Wildlife Trafficking in the Anthropocene: Conservation, Crime & Communities
Series: USC Center for Artificial Intelligence in Society (CAIS) Seminar Series
Abstract: Levels of unsustainable and illegal natural resource exploitation have escalated in scope, scale, and severity. Illegal over-harvest of plant and animal species occurs around the world and poses risks to species, ecosystems, and people. Beyond the risk of species loss, overexploitation represents stolen natural resources, is associated with corruption and insecurity, human rights abuses, and regional destabilization in some of the world's most vulnerable developing nations. This presentation will discuss conservation criminology-”an interdisciplinary and applied science for understanding risks to global natural resources.
Biography: Dr. Meredith Gore is a conservation social scientist leveraging concepts of risk to enhance understanding of human-environment relationships. Her scholarship is designed to build evidence for action. The majority of her scientific inquiry can be described as convergence research on conservation issues such as wildlife trafficking, illegal logging, fishing and mining. She received her PhD in Natural Resource Policy and Management from Cornell University, MA in Environment and Resource Policy from George Washington University, and BA in Anthropology and Environmental Studies from Brandeis University. She's a MSU Global Research Academy Fellow, National Academies of Sciences Jefferson Science Fellow, US Department of State Embassy Science Fellow and Emerging Wildlife Conservation Leader.
Host: USC Center for Artificial Intelligence in Society (CAIS)
Location: Ronald Tutor Hall of Engineering (RTH) - 211
Audiences: Everyone Is Invited
Contact: Computer Science Department
-
CS Colloquium: Ludwig Schmidt (UC Berkeley) - Do ImageNet Classifiers Generalize to ImageNet?
Thu, Mar 12, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Ludwig Schmidt, UC Berkeley
Talk Title: Do ImageNet Classifiers Generalize to ImageNet?
Series: CS Colloquium
Abstract: Progress on the ImageNet dataset seeded much of the excitement around the machine learning revolution of the past decade. In this talk, we analyze this progress in order to understand the obstacles blocking the path towards safe, dependable, and secure machine learning.
First, we will investigate the nature and extent of overfitting on ML benchmarks through reproducibility experiments for ImageNet and other key datasets. Our results show that overfitting through test set re-use is surprisingly absent, but distribution shift poses a major open problem for reliable ML.
In the second part, we will focus on a particular robustness issue, known as adversarial examples, and develop methods inspired by optimization and generalization theory to address this issue. We conclude with a large experimental study of current robustness interventions that summarizes the main challenges going forward.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Ludwig Schmidt is a postdoctoral researcher at UC Berkeley working with Moritz Hardt and Ben Recht. Ludwig's research interests revolve around the empirical and theoretical foundations of machine learning, often with a focus on making machine learning more reliable. Before Berkeley, Ludwig completed his PhD at MIT under the supervision of Piotr Indyk. Ludwig received a Google PhD fellowship, a Microsoft Simons fellowship, a best paper award at the International Conference on Machine Learning (ICML), and the Sprowls dissertation award from MIT.
Host: Haipeng Luo
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
**LOCATION CHANGE**CS Colloquium: Ioannis Panageas (SUTD) - Depth-width trade-offs for ReLU networks via Sharkovsky's theorem
Thu, Mar 12, 2020 @ 04:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Ioannis Panageas, Singapore University of Technology and Design
Talk Title: Depth-width trade-offs for ReLU networks via Sharkovsky's theorem
Series: CS Colloquium
Abstract: Understanding the representational power of Deep Neural Networks (DNNs) and how their structural properties (e.g., depth, width, type of activation unit) affect the functions they can compute, has been an important yet challenging question in deep learning and approximation theory. In a seminal paper, Telgarsky highlighted the benefits of depth by presenting a family of functions (based on simple triangular waves) for which DNNs achieve zero classification error, whereas shallow networks with fewer than exponentially many nodes incur constant error. Even though Telgarsky's work reveals the limitations of shallow neural networks, it does not inform us on why these functions are difficult to represent and in fact he states it as a tantalizing open question to characterize those functions that cannot be well-approximated by smaller depths. In this talk, we will point to a new connection between DNNs expressivity and Sharkovsky's Theorem from dynamical systems, that enables us to characterize the depth-width trade-offs of ReLU networks for representing functions based on the presence of generalized notion of fixed points, called periodic points (a fixed point is a point of period 1). Motivated by our observation that the triangle waves used in Telgarsky's work contain points of period 3 - a period that is special in that it implies chaotic behavior based on the celebrated result by Li-Yorke - we will give general lower bounds for the width needed to represent periodic functions as a function of the depth. Technically, the crux of our approach is based on an eigenvalue analysis of the dynamical system associated with such functions.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Ioannis Panageas is an Assistant Professor at Information Systems Department of SUTD since September 2018. Prior to that he was a MIT postdoctoral fellow working with Constantinos Daskalakis. He received his PhD in Algorithms, Combinatorics and Optimization from Georgia Institute of Technology in 2016, a Diploma in EECS from National Technical University of Athens (summa cum laude) and a M.Sc. in Mathematics from Georgia Institute of Technology. His work lies on the intersection of optimization, probability, learning theory, dynamical systems and algorithms. He is the recipient of the 2019 NRF fellowship for AI (analogue of NSF CAREER award).
URL Website: https://panageas.github.io/
Host: Shaddin Dughmi
Location: Ronald Tutor Hall of Engineering (RTH) - 115
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
Seminar will be exclusively online (no in-room presentation) - CS Colloquium: Simon S. Du (Princeton University) - Foundations of Learning Systems with (Deep) Function Approximators
Tue, Mar 24, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Simon S. Du, Princeton University
Talk Title: Foundations of Learning Systems with (Deep) Function Approximators
Series: CS Colloquium
Abstract: Function approximators, such as deep neural networks, play a crucial role in building learning systems that make predictions and decisions. In this talk, I will discuss my work on understanding, designing, and applying function approximators.
First, I will focus on understanding deep neural networks. The main result is that the over-parameterized neural network is equivalent to a new kernel, Neural Tangent Kernel. This equivalence implies two surprising phenomena: 1) the simple algorithm gradient descent provably finds the global optimum of the highly non-convex empirical risk, and 2) the learned neural network generalizes well despite being highly over-parameterized. Furthermore, this equivalence helps us design a new class of function approximators: we transform (fully-connected and graph) neural networks to (fully-connected and graph) Neural Tangent Kernels, which achieve superior performance on standard benchmarks.
In the second part of the talk, I will focus on applying function approximators to decision-making, aka reinforcement learning, problems. In sharp contrast to the (simpler) supervised prediction problems, solving reinforcement learning problems requires an exponential number of samples, even if one applies function approximators. I will then discuss what additional structures that permit statistically efficient algorithms.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Simon S. Du is a postdoc at the Institute for Advanced Study of Princeton, hosted by Sanjeev Arora. He completed his Ph.D. in Machine Learning at Carnegie Mellon University, where he was co-advised by Aarti Singh and Barnabás Póczos. Previously, he studied EECS and EMS at UC Berkeley. He has also spent time at Simons Institute and research labs of Facebook, Google, and Microsoft. His research interests are broadly in machine learning, with a focus on the foundations of deep learning and reinforcement learning.
Host: Haipeng Luo
Location: Seminar will be exclusively online (no in-room presentation)
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
Computer Science General Faculty Meeting
Wed, Mar 25, 2020 @ 12:00 PM - 02:00 PM
Thomas Lord Department of Computer Science
Receptions & Special Events
Bi-Weekly regular faculty meeting for invited full-time Computer Science faculty only. Event details emailed directly to attendees.
Audiences: Invited Faculty Only
Contact: Assistant to CS chair
-
Seminar will be exclusively online (no in-room presentation) - CS Colloquium: Farnaz Behrang (Georgia Institute of Technology) - Leveraging Existing Software Artifacts to Support Design, Development, and Testing of Mobile Applications
Wed, Mar 25, 2020 @ 02:00 PM - 03:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Farnaz Behrang, Georgia Institute of Technology
Talk Title: Leveraging Existing Software Artifacts to Support Design, Development, and Testing of Mobile Applications
Series: CS Colloquium
Abstract: We are living in the era of big data, in which generating and sharing data has become much easier, and massive amounts of information are created in a fraction of a second. In the context of software engineering, in particular, the number of open-source software repositories (e.g., GitHub, Bitbucket, SourceForge) where software developers share their software artifacts is ever-increasing, and hundreds of millions of lines of code are freely available and easily accessible. This has resulted in an increasing interest in analyzing the rich data available in such repositories. In the past decade, researchers have been mining online repositories to take advantage of existing source code to support different development activities, such as bug prediction, refactoring, and API updates. Despite the large number of proposed techniques that leverage existing source code, however, these techniques mostly focus on supporting coding activities. Other important software engineering tasks, such as software design and testing, have been mostly ignored by previous work.
In this talk, I will present my research on leveraging existing source code and other related artifacts (e.g., test cases) to support the design, development, and testing of mobile applications using automated techniques. I will first present a technique that leverages the growing number of open-source apps in public repositories to support app design and development. I will then present techniques that take advantage of existing test cases to reduce the cost of testing mobile apps. I will conclude my talk sketching future research directions that I plan to pursue.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Farnaz Behrang is a Ph.D. candidate in the School of Computer Science at the Georgia Institute of Technology. Her research interests lie primarily in the area of software engineering, with a focus on software analysis and testing. Her research goal is to develop automated techniques and tools that improve software quality and developer productivity. Her work has been recognized with several awards including ACM SIGSOFT Distinguished Paper Awards at MOBILESOFT 2018 and FSE 2015.
Host: Chao Wang
Location: Seminar will be exclusively online (no in-room presentation)
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
Seminar will be exclusively online (no in-room presentation) - CS Colloquium: Aditya Grover (Stanford University) - Machine Learning for Accelerating Scientific Discovery
Thu, Mar 26, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Aditya Grover, Stanford University
Talk Title: Machine Learning for Accelerating Scientific Discovery
Series: CS Colloquium
Abstract: The dramatic increase in both sensor capabilities and computational power over the last few decades has created enormous opportunities for using machine learning (ML) to enhance scientific discovery. To realize this potential, ML systems must seamlessly integrate with the key tools for scientific discovery. For instance, how can we incorporate scientific domain knowledge within ML algorithms? How can we use ML to quantify uncertainty in simulations? How can we use ML to plan experiments under real-world budget constraints? For these questions, I'll first present the key computational and statistical challenges through the lens of probabilistic modeling. Next, I'll highlight limitations of existing approaches for scaling to high-dimensional data and present algorithms from my research that can effectively overcome these challenges. These algorithms are theoretically principled, domain-agnostic, and exhibit strong empirical performance. Notably, I'll describe a collaboration with chemists and material scientists where we used probabilistic models to efficiently optimize an experimental pipeline for electric batteries. Finally, I'll conclude with an overview of future opportunities for using ML to accelerate scientific discovery.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Aditya Grover is a fifth-year Ph.D. candidate in Computer Science at Stanford University advised by Stefano Ermon. His research focuses on probabilistic modeling and reasoning and is grounded in real-world scientific applications. Aditya's research has been published in top scientific and ML/AI venues (e.g., Nature, NeurIPS, ICML, ICLR, AAAI, AISTATS), included in widely-used open source ML software, and deployed into production at major technology companies. His work has been recognized with a best paper award (StarAI), a Lieberman Fellowship, a Data Science Institute Scholarship, and a Microsoft Research Ph.D. Fellowship. He is also a Teaching Fellow at Stanford since 2018, where he co-created and teaches a new class on Deep Generative Models. Previously, Aditya obtained his bachelors in Computer Science and Engineering from IIT Delhi in 2015, where he received a best undergraduate thesis award.
Host: Bistra Dilkina
Location: Seminar will be exclusively online (no in-room presentation)
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
Seminar will be exclusively online (no in-room presentation) - CS Colloquium: Zhihao Jia (Stanford University) - Automated Discovery of Machine Learning Optimizations
Thu, Mar 26, 2020 @ 04:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Zhihao Jia, Stanford University
Talk Title: Automated Discovery of Machine Learning Optimizations
Series: CS Colloquium
Abstract: As an increasingly important workload, machine learning (ML) applications require different performance optimization techniques from traditional runtimes and compilers. In particular, to accelerate ML applications, it is generally necessary to perform ML computations on heterogeneous hardware and parallelize computations using multiple data dimensions, neither of which is even expressible in traditional compilers and runtimes. In this talk, I will describe my work on automated discovery of performance optimizations to accelerate ML computations.
TASO, the Tensor Algebra SuperOptimizer, optimizes the computation graphs of deep neural networks (DNNs) by automatically generating potential graph optimizations and formally verifying their correctness. TASO outperforms rule-based graph optimizers in existing ML systems (e.g., TensorFlow, TensorRT, and TVM) by up to 3x by automatically discovering novel graph optimizations, while also requiring significantly less human effort.
FlexFlow is a system for accelerating distributed DNN training. FlexFlow identifies parallelization dimensions not considered in existing ML systems (e.g., TensorFlow and PyTorch) and automatically discovers fast parallelization strategies for a specific parallel machine. Companies and national labs are using FlexFlow to train production ML models that do not scale well in current ML systems, achieving over 10x performance improvement.
I will also outline future research directions for further automating ML systems, such as codesigning ML models, software systems, and hardware backends for end-to-end ML deployment.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Zhihao Jia is a Ph.D. candidate in the Computer Science department at Stanford University working with Alex Aiken and Matei Zaharia. His research interests lie in the intersection of computer systems and machine learning, with a focus on building efficient, scalable, and high-performance systems for ML computations.
Host: Leana Golubchik
Location: Seminar will be exclusively online (no in-room presentation)
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
Seminar will be exclusively online (no in-room presentation) - CS Colloquium: Alan Liu (Carnegie Mellon University) - Enabling Future-Proof Telemetry for Networked Systems
Tue, Mar 31, 2020 @ 11:00 AM - 12:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Alan Liu, Carnegie Mellon University
Talk Title: Enabling Future-Proof Telemetry for Networked Systems
Series: CS Colloquium
Abstract: Today's networked systems, such as data center, cellular, and sensor networks, face increasing demands on security, performance, and reliability. To fulfill these demands, we first need to obtain timely and accurate telemetry information about what is happening in the system. For instance, understanding the volume and the number of distinct network connections can help detect and mitigate network attacks. In storage systems, identifying hot items can help balance the server load. Unfortunately, existing telemetry tools cannot robustly handle multiple telemetry tasks with diverse workloads and resource constraints.
In this talk, I will present my research that focuses on building telemetry systems that are future-proof for current and unforeseen telemetry tasks, diverse workloads, and heterogeneous platforms. I will discuss the efficient algorithms and implementations that realize this future-proof vision in network monitoring for hardware and software platforms. I will describe how bridging theory and practice with sketching and sampling algorithms can significantly reduce memory footprints and speedup computations while providing robust results. Finally, I will end the talk with new directions in obtaining future-proof analytics for other types of networked systems, such as low-power sensors and mobile devices, while enhancing energy efficiency and data privacy.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Alan (Zaoxing) Liu is a postdoctoral researcher at Carnegie Mellon University. His research interests are in networked and distributed systems with a recent focus on efficient system and algorithmic design for telemetry, big-data analytics, and privacy. His research papers have been published in venues such as ACM SIGCOMM, USENIX FAST, and OSDI. He is a recipient of the best paper award at USENIX FAST'19 for his work on large-scale distributed load balancing. His work received multiples recognitions, including ACM STOC "Best-of-Theory" plenary talk and USENIX ATC "Best-of-Rest". Prior to CMU, he obtained his Ph.D. in Computer Science from Johns Hopkins University.
Host: Ramesh Govindan
Location: Seminar will be exclusively online (no in-room presentation)
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
Seminar will be exclusively online (no in-room presentation) - CS Colloquium: Baharan Mirzasoleiman (Stanford University) - Efficient Machine Learning via Data Summarization
Tue, Mar 31, 2020 @ 04:00 PM - 05:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Baharan Mirzasoleiman, Stanford University
Talk Title: Efficient Machine Learning via Data Summarization
Series: CS Colloquium
Abstract: Large datasets have been crucial to the success of modern machine learning models. However, training on massive data has two major limitations. First, it is contingent on exceptionally large and expensive computational resources, and incurs a substantial cost due to the significant energy consumption.
Second, in many real-world applications such as medical diagnosis and self-driving cars, big data contains highly imbalanced classes and noisy labels. In such cases, training on the entire data does not result in a high-quality model. In this talk, I will argue that we can address the above limitations by developing techniques that can identify and extract the representative subsets from massive datasets. Training on representative subsets not only reduces the substantial costs of learning from big data, but also improves their accuracy and robustness against noisy labels. I will present two key aspects to achieve this goal: (1) extracting the representative data points by summarizing massive datasets; and (2) developing efficient optimization methods to learn from the extracted summaries. I will discuss how we can develop theoretically rigorous techniques that provide strong guarantees for the quality of the extracted summaries, and the learned models' quality and robustness against noisy labels. I will also show the applications of these techniques to several problems, including summarizing massive image collections, online video summarization, and speeding up training machine learning models.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Biography: Baharan Mirzasoleiman is a Postdoctoral Research Scholar in Computer Science Department at Stanford University, where she works with Prof. Jure Leskovec. Baharan's research focuses on developing new methods that enable efficient exploration and learning from massive datasets. She received her PhD from ETH Zurich, working with Prof. Andreas Krause. She has also spent two summers as an intern at Google Research. She was awarded an ETH medal for Outstanding Doctoral Dissertation, and a Google Anita Borg Memorial Scholarship. She was also selected as a Rising Star in EECS from MIT.
Host: Bistra Dilkina
Location: Seminar will be exclusively online (no in-room presentation)
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
