- talks - lists.cs.princeton.edu

William Pierson Field Lecture: Roberto Martin-Martin, Mon Dec 5 at 1:30pm
by Emily C. Lawrence 29 Nov '22

29 Nov '22

William Pierson Field Lecture Speaker: Roberto Martin-Martin, University of Texas at Austin Time: 1:30pm - 2:50pm, Monday Dec 5, 2022 Location: CS 104 Title: What do we want robots to do for us (and how could we get there)? — Challenges in Robot Perception for Long Horizon Heterogeneous Tasks Abstract:: Latest advances in robot learning are creating more capable and robust embodied agents, making us consider a future where robots could help humans in their everyday lives. But, what exactly do we want robots to do for us? In this talk, I will consider this question and discuss the results of a first analysis: a new robot benchmark with a human-guided set of tasks in ecological household environments. This benchmark demands solutions with better capabilities for long-horizon planning and control; I will present our first steps in robot perception towards improved planning and control: perceiving human plans to long-horizon tasks from instructional videos, and predicting video to enable versatile visuomotor control. Bio: Roberto Martin-Martin is Assistant Professor of Computer Science at University of Texas at Austin. His research connects robotics, computer vision and machine learning. He studies and develops novel AI algorithms that enable robots to perform tasks in human uncontrolled environments such as homes and offices. He creates novel decision-making solutions based on reinforcement learning, imitation learning, planning and control, and explores topics in robot-perception such as pose estimation and tracking, video prediction and parsing. Martin-Martin received his Ph.D. from Berlin Institute of Technology (TUB) prior to a postdoctoral position at the Stanford Vision and Learning Lab. His work has been selected as RSS Best Systems Paper Award, RSS Pioneer, Winner of the Amazon Picking Challenge, and ICRA and IROS Best Paper Nominee. He is chair of the IEEE Technical Committee in Mobile Manipulation.

1 0

Tyler Park (QCB) Dissertation Defense: Dec 7th at 10am
by Gradinfo 28 Nov '22

28 Nov '22

The Lewis-Sigler Institute for Integrative Genomics ~ Graduate Program in Quantitative and Computational Biology Final Public Oral Tyler Park Integrating multi-modal biological data uncovers vulnerabilities and complex interactions in cancer Wednesday, December 7, 2022 10:00 AM Computer science building, room 301 Raphael Lab A copy of the dissertation is available to QCB for inspection electronically; email [ mailto:jgiraldi@princeton.edu | jgiraldi(a)princeton.edu ] with your request.

1 0

Meryem Essaidi will present her FPO "User-Centered Algorithmic Mechanism Design" on Tuesday, December 20, 2022 at 1pm via Zoom.
by Nicki Mahler 23 Nov '22

23 Nov '22

Meryem Essaidi will present her FPO "User-Centered Algorithmic Mechanism Design" on Tuesday, December 20, 2022 at 1pm via Zoom. Zoom link: [ https://princeton.zoom.us/j/93585578182 | https://princeton.zoom.us/j/93585578182 ] Committee: Adviser: Matt Weinberg Readers: Sam Taggart, Mark Braverman Examiners: Olga Russakovsky, Huacheng Yu All are welcome to attend. Abstract: In algorithmic mechanism design, classical desiderata define standards used to evaluate the performance of an algorithm or mechanism. Examples of such desiderata are efficiency, revenue maximization, and strategyproofness. These desiderata are oftentimes seller-centered and tend to create negative externalities that harm the users targeted by the market. Examples of such externalities are not treating the users equally, not protecting against adversarial sellers, and not maximizing user utility. In this thesis, we explore the frontier between user-centric and seller-centric performance at different levels of leverage offered by central regulation. We ask: how can we design new solutions that prioritize user-centered desiderata while maintaining the pre-existing seller-centered desiderata or without too great a cost to them? We present theoretical frameworks at three levels of intervention“protected”goods, and significant regulation for essential and public goods. Our first framework is credible auctions. We consider a revenue-maximizing seller with a single item for sale to multiple users with independent-and-identicallydistributed valuations. In this work, assuming the existence of cryptographically-secure commitment-schemes, we identify a new single-item auction that is credible, strategyproof, revenue-optimal, and terminates in constant rounds in expectation for all distributions with finite monopoly price. Our second framework is fair and symmetric ad auctions. We consider a revenuemaximizing seller with multiple items for sale to a single population of additive buyers with independent item values. We motivate this via fairness in ad auctions where items correspond to (views from) users, and equally-qualified users from different demographics should be shown the same desired ad at equal rates. We show that bundling all items together achieves a constant-factor approximation to the revenue-optimal itemsymmetric mechanism; as any item-symmetric auction is also fair. Observe that in this domain, bundling all items together corresponds to concealing all demographic data. i Our last framework is markets with mandatory purchase. We study a problem inspired by regulated health-insurance markets, and investigate whether limiting entry of providers increases or decreases the utility(welfare minus revenue) of users who purchase from the providers, specifically in settings where the outside option of “purchasing nothing” is prohibitively undesirable.

2 1

CSML Lunch and Learn: Robustness for Models and Algorithms in Machine Learning, Monday, Nov 28
by Emily C. Lawrence 23 Nov '22

23 Nov '22

CSML Lunch and Learn SEMINAR Speaker: Yassine Laguel, Rutgers University Title: Robustness for Models and Algorithms in Machine Learning Day: Monday, November 28, 2022 Time: 12:30 p.m. Lunch available from 12:00 p.m. Location: CSML, 26 Prospect Ave. Classroom 103 Host: CSML and Mert Gürbüzbalaban Abstract: Risk-averse optimization plays a major role in the design of safety for machine learning applications. In this talk, we will present a set of tools to enhance the robustness of models and algorithms to potentially harmful data shifts. First, we will expose some general results on the modeling of risk aversion and highlight the interest of superquantile-based risk measures to enforce robustness to worst-case events. We will then show how such a measure may be minimized in the centralized setting based on a smoothing à la Nesterov of the superquantile. We will then present applications of this framework in federated learning to handle statistical heterogeneity among the devices of a given network. Second, we will revisit the bias-variance trade-off of first-order stochastic algorithms from a robust perspective. Precisely, we study the convergence properties of accelerated methods on saddle-point problems for diverse robustness metrics. We deduce new ways to set the associated hyperparameters to stabilize the algorithm at equilibrium. Bio : Yassine Laguel is a postdoctoral researcher at Rutgers University and a departmental guest at the Center for Statistics and Machine Learning at Princeton University. He received his M.Sc. in applied mathematics from the Ecole Nationale Supérieure de l’Informatique et des Mathématiques Appliquées (ENSIMAG) in 2018. He then completed his Ph.D. under the supervision of Jérôme Malick at the Université Grenoble Alpes. He is now working in the Management Sciences and Information Systems department at Rutgers Business School, in the team of Mert Gürbüzbalaban. Dr. Laguel’s research interests center around optimization under uncertainty and its applications to stochastic programming and machine learning. A common thread in his research is the design and analysis of numerical algorithms to address risk in data-driven applications.

1 0

Andrew Jones will present his FPO "Probabilistic models for structured biomedical data" on Friday, December 16, 2022 at 9:30 AM in COS 402 and Zoom.
by gradinfo＠cs.princeton.edu 23 Nov '22

23 Nov '22

Andrew Jones will present his FPO "Probabilistic models for structured biomedical data" on Friday, December 16, 2022 at 9:30 AM in COS 402 and Zoom. Location: Zoom link: https://princeton.zoom.us/j/95479201507 <https://www.google.com/url?q=https://princeton.zoom.us/j/95479201507&sa=D&s ource=calendar&ust=1669648765658473&usg=AOvVaw0W0kAo0lJDADvaWIY_Kr4D> The members of Andrew's committee are as follows: Examiners: Barbara Engelhardt (Adviser), Ben Raphael, Adji Bousso Dieng Readers: Jonathan Pillow, Olga Russakovsky A copy of his thesis will be available, upon request, two weeks before the FPO. Please email gradinfo(a)cs.princeton.edu <mailto:gradinfo@cs.princeton.edu> if you would like a copy of the thesis. Everyone is invited to attend his talk. Abstract follows below: Modern biomedical datasets-from molecular measurements of gene expression to pathology images-hold promise for discovering new therapeutics and probing basic questions about the behavior of cells. Thoughtful statistical modeling of these complex, high-dimensional data is crucial to elucidate robust scientific findings. A common assumption in data analysis that the data samples are independent and identically distributed. However, this assumption is nearly always violated in practice. This is especially true in the setting of biomedical data, which often exhibit some amount of structure, such as subgroups of patients, cells, or tissue types or other correlation structure among the samples. In this body of work, I propose data analysis and experimental design frameworks to account for several types of highly-structured biomedical data. These approaches, which take the form of Bayesian models and associated inference algorithms, are specifically tailored for datasets with group structure, multiple data modalities, and spatial organization of samples. In the first line of work, I propose a model for contrastive dimension reduction that decomposes the sources of variation in samples that belong to case and control conditions. Second, I propose a computational framework for aligning spatially-resolved genomics data into a common coordinate system that accounts for spatial correlation among the samples and models multiple data modalities. Finally, I propose a family of methods for optimally designing spatially-resolved genomics experiments that is tailored to the highly-structured data collection process of these studies. Together, this body of work advances the field of biomedical data analysis by developing models that directly exploit common types of structure within these data and demonstrating the advantage of these modeling approaches across an array of data types.

1 0

CS Colloquium: Shuran Song, Monday, Nov 7 at 12:30pm
by Emily C. Lawrence 21 Nov '22

21 Nov '22

CS Colloquium Series Speaker: Shuran Song, Columbia University Date: Monday, November 7 Time: 12:30pm Location: CS 105 Host: Szymon Rusinkiewicz Event page: [ https://www.cs.princeton.edu/events/26279 | https://www.cs.princeton.edu/events/26279 ] Title: Unpatched Vulnerabilities in Cellular Standards Abstract: Despite the incredible capabilities (speed, repeatability) of our hardware, most robot manipulators today are deliberately programmed to avoid dynamics – moving slow enough so they can adhere to quasi-static assumptions about the world. In contrast, people frequently (and subconsciously) make use of dynamic phenomena to manipulate everyday objects – from unfurling blankets to tossing trash, to improve efficiency and physical reach range. These abilities are made possible by an intuition of physics, a cornerstone of intelligence. How do we impart the same on robots? In this talk, I will discuss how we might enable robots to leverage dynamics for manipulation in unstructured environments. Modeling the complex dynamics of unseen objects from pixels is challenging. However, by tightly integrating perception and action, we show it is possible to relax the need for accurate dynamical models. Thereby, allowing robots to (i) learn dynamic skills for complex objects, (ii) adapt to new scenarios using visual feedback, and (iii) use their dynamic interactions to improve their understanding of the world. By changing the way we think about dynamics – from avoiding it to embracing it – we can simplify a number of classically challenging problems, leading to new robot capabilities. Bio: Shuran Song is an Assistant Professor in the Department of Computer Science at Columbia University. Before that, she received her Ph.D. in Computer Science at Princeton University, BEng. at HKUST. Her research interests lie at the intersection of computer vision and robotics. Song’s research has been recognized through several awards including the Best Paper Awards at RSS’22 and T-RO’20, Best System Paper Awards at CoRL’21, RSS’19, and Amazon Robotics’18, and finalist at RSS’22, ICRA'20, CVPR'19, RSS’19, and IROS'18. She is also a recipient of the NSF Career Award, as well as research awards from Microsoft, Toyota Research, Google, Amazon, JP Morgan, and Sloan Foundation. To learn more about Shuran’s work please visit: [ https://www.cs.columbia.edu/~shurans/ | https://www.cs.columbia.edu/~shurans/ ]

1 12

Claudia Roberts will present her FPO "Human-machine Collaboration in Real-World Machine-Learning Applications” on Thursday, December 15, 2022 at 2:00pm in CS 105 (correction)
by Nicki Mahler 18 Nov '22

18 Nov '22

Claudia Roberts will present her FPO "Human-machine Collaboration in Real-World Machine-Learning Applications” on Thursday, December 15, 2022 at 2:00pm in CS 105. The members of her committee are as follows: Arvind Narayanan (Adviser); Adji Dieng and Barbara Engelhardt (Readers); Andrés Monroy-Hernández and Matt Salganik (Examiners) All are welcome to attend. Abstract: Automation tools like machine learning are a necessity in our big data world. Thanks to the Internet and advancements in all facets of computer and storage technology, almost everyone has a voice in the Internet connected world. However, there are still very real physical limits in our physical world. This dichotomy—the seemingly limitless nature of technology enabled data colliding with the physical limits of the real world—has made automation tools a necessity, and predictive models powered by machine learning algorithms are one such tool. The promise of machine learning to accurately predict future human behavior and human preferences has lead practitioners and researchers alike to apply machine learning automation tools to tasks such as product recommendations and speculatory activities such as long term job applicant success. However, due to the mercurial nature of humans, developing mathematical intermediaries to attempt to model and predict human behavior is challenging and not a straight-forward task. One way of harnessing the power of machine-learning backed automation to help reduce the scale of many real-world applications in more challenging domain settings is by having humans and machines collaborating in non-trivial ways. In this dissertation, we delineate the various ways in which humans and machines collaborate in challenging real-world applications. Moreover, we highlight three specific ways in which we can use human-machine collaboration to keep or increase utility and reduce real-world harm when using these systems in the wild: (i) humans enabling computers with domain specific knowledge, (ii) computers providing humans with algorithmic explanations, (iii) humans and computers working together in decision making. [ https://www.cs.princeton.edu/events/26295 | https://www.cs.princeton.edu/events/26295 ]

1 0

Claudia Roberts will present her FPO "Human-machine Cooperation in Machine-Learning Resistant Real-World Applications" on Thursday, December 15, 2022 at 2:00pm in CS 105.
by Nicki Mahler 17 Nov '22

17 Nov '22

Claudia Roberts will present her FPO "Human-machine Cooperation in Machine-Learning Resistant Real-World Applications" on Thursday, December 15, 2022 at 2:00pm in CS 105. The members of her committee are as follows: Arvind Narayanan (Adviser); Adji Dieng and Barbara Engerhardt (Readers); Andrés Monroy-Hernández and Matt Salganik (Examiners) All are welcome to attend. Abstract: Automation tools like machine learning are a necessity in our big data world. Thanks to the Internet and advancements in all facets of computer and storage technology, almost everyone has a voice in the Internet connected world. However, there are still very real physical limits in our physical world. This dichotomy—the seemingly limitless nature of technology enabled data colliding with the physical limits of the real world—has made automation tools a necessity, and predictive models powered by machine learning algorithms are one such tool. The promise of machine learning to accurately predict future human behavior and human preferences has lead practitioners and researchers alike to apply machine learning automation tools to tasks such as product recommendations and speculatory activities such as long term job applicant success. However, due to the mercurial nature of humans, developing mathematical intermediaries to attempt to model and predict human behavior is challenging and not a straight-forward task. One way of harnessing the power of machine-learning backed automation to help reduce the scale of many real-world applications in more challenging domain settings is by having humans and machines collaborating in non-trivial ways. In this dissertation, we delineate the various ways in which humans and machines collaborate in challenging real-world applications. Moreover, we highlight three specific ways in which we can use human-machine collaboration to keep or increase utility and reduce real-world harm when using these systems in the wild: (i) humans enabling computers with domain specific knowledge, (ii) computers providing humans with algorithmic explanations, (iii) humans and computers working together in decision making. [ https://www.cs.princeton.edu/events/26295 | https://www.cs.princeton.edu/events/26295 ]

1 0

David Liu will present his FPO "A Serverless Architecture for Application-level Orchestration" on Wednesday, December 14, 2022 at 2pm in CS 402
by Nicki Mahler 17 Nov '22

17 Nov '22

David Liu will present his FPO "A Serverless Architecture for Application-level Orchestration" on Wednesday, December 14, 2022 at 2pm in CS 402. The members of his committee is as follows: Amit Levy and Mike Freedman (Advisers); Wyatt Lloyd and Ravi Netravali (Readers); Mike Freedman, Jennifer Rexford (Examiners) All are welcome to attend. Abstract: This thesis examines the problem of building large-scale applications using the serverless computing model and proposes decentralized, application-level orchestration for serverless workloads. Comparing with standalone orchestrators, the state-of-the-art solution to building large-scale serverless applications, we demonstrate that application-level orchestration is possible and practical using just the basic APIs of existing serverless infrastructures, and that it benefits both cloud users and cloud providers. It empowers cloud users with the flexibility of application-specific optimizations. It frees cloud providers from hosting and maintaining yet another performance-critical service. Furthermore, the performance and efficiency of application-level orchestration improve as the underlying systems develop. Thus, cloud providers can direct freed-up resources to core services in their serverless infrastructure and automatically reaps the benefits of a better orchestrator. This thesis describes mechanisms and implementations that help realize the goal of applicationlevel orchestration. In particular, we explain the necessity and challenges of decentralizing orchestration, and present a system for decentralized orchestration named Unum. Unum introduces an intermediate representation (IR) language to express execution graphs using only node-local information to decentralize the orchestration logic of applications. Unum implements orchestration as a library that runs in-situ with user-defined FaaS functions, rather than as a standalone service. The library relies on a minimal set of existing serverless APIs—function invocation and a few basic datastore operations—that are common across cloud platforms. Unum ensures workflow correctness despite multiple executions of non-deterministic functions by using checkpoints to commit to exactly one output for a function invocation. Our results show that a representative set of applications scale better, run faster, and cost significantly less with Unum than a state-of-the-art centralized orchestrator. We also show that Unum’s IR allows hand-tuned applications to run faster by using application-specific optimizations and supporting a richer set of application patterns. We hope the results in this thesis inspire cloud practitioners to reconsider the approach of supporting new functionalities by simply adding more services to the cloud infrastructure. And we hope to encourage the building of other application-level orchestration systems from the serverless community. [ https://www.cs.princeton.edu/events/26287 | https://www.cs.princeton.edu/events/26287 ]

1 1

ECE Seminar: Jakub Szefer, Friday, Nov 18 at 12:30 pm
by Emily C. Lawrence 16 Nov '22

16 Nov '22

ECE Seminar Speaker: Jakub Szefer, Associate Professor, Dept. of Electrical Engineering, Yale University Day: November 18, 2022 Time: 12:30 pm EST Location: EQuad B205 Host: David Wentzlaff Title: Quantum Computer Cybersecurity Abstract: Quantum computers are becoming larger every year, and are already today easily accessible thanks to different cloud-based services from companies such as IBM Quantum, Microsoft Azure, or Amazon Braket. Increasing quantum computer sizes and computational abilities bring closer the potential for these computers to be able to generate novel solutions and results in chemistry, medicine, or material science, for example. However, as these computers promise to generate novel results and data, they will become targets of cyber attackers, who may want to steal information, influence the computation, or cause disruption to the operation of the quantum computers. The security threat-level is heightened by the fact that these computers are network-connected and are open to untrusted cloud users running malicious code on them. In this talk, we will present some of the first research on security evaluation of real, cloud-based quantum computers. We will first discuss potential “software” threats to quantum computers, by demonstrating insecurity of some of the basic gates, or operations, that are today available on the quantum computers. We will also demonstrate potential security threats considering future, multi-tenant use of the quantum computers where multiple users share a single quantum computer. The potential threats were found and evaluated by collecting data and running quantum programs on real, cloud-based quantum computers. The talk will also touch upon defenses we have proposed. In addition, the talk will aim to motivate the need to study cybersecurity of quantum computers and include security protections in their design. Today, large amounts of money and effort are put into development of bigger and more capable quantum computers, quantum software frameworks, and cloud infrastructures for these computers. Meanwhile, their security analysis and protections are lagging much behind. Bio: Prof. Jakub Szefer is currently an Associate Professor of Electrical Engineering at Yale University, where he leads the Computer Architecture and Security Laboratory (CASLAB). Jakub’s research encompasses secure processor architectures, cloud and device security, fast implementation of post-quantum cryptographic algorithms, as well as quantum computer cybersecurity. Among others, Jakub is the author of first book focusing on processor architecture security: “Principles of Secure Processor Architecture Design”, published in 2018. Most recently, Jakub is a recipient of the 2021 Ackerman Award for Teaching and Mentoring from Yale University.

1 0