Select a calendar:
Filter March Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Conferences, Lectures, & Seminars
Events for March
-
CS Colloquium: Sergey Levine (UC Berkeley) - Deep Learning for Decision Making and Control
Tue, Mar 03, 2015 @ 09:45 AM - 10:50 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Sergey Levine, UC Berkeley
Talk Title: Deep Learning for Decision Making and Control
Series: CS Colloquium
Abstract: A remarkable feature of human and animal intelligence is the ability to autonomously acquire new behaviors. My work is concerned with designing algorithms that aim to bring this ability to robots and simulated characters. A central challenge in this field is to learn behaviors with representations that are sufficiently general and expressive to handle the wide range of motion skills that are necessary for real-world applications, such as general-purpose household robots. These representations must also be able to operate on raw, high-dimensional inputs and outputs, such as camera images, joint torques, and muscle activations. I will describe a class of guided policy search algorithms that tackle this challenge by transforming the task of learning control policies into a supervised learning problem, with supervision provided by simple, efficient trajectory-centric methods. I will show how this approach can be applied to a wide range of tasks, from locomotion and push recovery to robotic manipulation. I will also present new results on using deep convolutional neural networks to directly learn policies that combine visual perception and control, learning the entire mapping from rich visual stimuli to motor torques on a real robot. I will conclude by discussing future directions in deep sensorimotor learning and how advances in this emerging field can be applied to a range of other areas.
The lecture will be streamed through the dedicated link HERE.
Biography: Sergey Levine is a postdoctoral researcher working with Professor Pieter Abbeel at UC Berkeley. He completed his PhD in 2014 with Vladlen Koltun at Stanford University. His research focuses on robotics, machine learning, and computer graphics. In his PhD thesis, he developed a novel guided policy search algorithm for learning rich, expressive locomotion policies. In later work, this method enabled learning a range of robotic manipulation tasks, as well as end-to-end training of policies for perception and control. He has also developed algorithms for learning from demonstration, inverse reinforcement learning, and data-driven character animation.
Host: Computer Science Department
More Info: https://bluejeans.com/658994068
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
Event Link: https://bluejeans.com/658994068
-
CS Colloquium: Prof. Ameet Talwalkar (UCLA) - Scalable and User-Friendly Machine Learning in Apache Spark
Thu, Mar 05, 2015 @ 04:00 PM - 05:15 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Ameet Talwalkar, UCLA
Talk Title: Scalable and User-Friendly Machine Learning in Apache Spark
Series: CS Colloquium
Abstract: Modern datasets are rapidly growing in size and complexity, and this wealth of data holds the promise for many transformational applications. Machine learning is seemingly poised to deliver on this promise, having proposed and rigorously evaluated a wide range of data processing techniques over the past several decades. However, concerns over scalability and usability present major roadblocks to the wider adoption of these methods. In this talk I will describe the MLbase project, which aims to address these concerns by developing machine learning functionality on top of Apache Spark, a popular cluster computing engine designed for iterative computation. I will first describe MLlib, Sparkâs scalable machine learning library that grew out of the MLbase project. I will also discuss higher level components of MLbase, focusing on the problem of hyperparameter optimization as a means to simplify the task of machine learning pipeline construction.
Biography: Ameet Talwalkar is an assistant professor of Computer Science at UCLA and a technical advisor for Databricks. His research addresses scalability and ease-of-use issues in the field of statistical machine learning, with applications in computational genomics. He led the initial development of the MLlib project in Apache Spark and is a co-author of the graduate-level textbook 'Foundations of Machine Learning' (2012, MIT Press). Prior to UCLA, he was an NSF post-doctoral fellow in the AMPLab at UC Berkeley. He obtained a B.S. from Yale University and a Ph.D. from the Courant Institute at NYU.
Host: Fei Sha
Location: Henry Salvatori Computer Science Center (SAL) - 101
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Elette Boyle (Technion Israel Institute of Technology) - Large-Scale Secure Computation
Tue, Mar 10, 2015 @ 09:45 AM - 10:50 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Elette Boyle, Technion Israel Institute of Technology
Talk Title: Large-Scale Secure Computation
Series: CS Colloquium
Abstract: The ability to collect and process large data sets has unlocked exciting new technological and research discoveries. Unfortunately, in several important applications, it is not possible to leverage the full extent of collected data, when information consists of sensitive data sets held by individual agents who are either unable or unwilling to share the data itself (e.g., patients' medical information gathered by different medical studies).
A promising approach to enable data sharing within these scenarios is to make use of cryptographic tools such as secure multi-party computation (MPC). MPC protocols provide a means for mutually untrusting parties to jointly evaluate a global function f over their secret inputs, while guaranteeing that no information is revealed beyond the function output.
However, despite great progress in MPC techniques in the last three decades, the surrounding world of data aggregation and computation has leapt even more rapidly forward. For example, nearly all existing MPC protocols require each party to store information comparable to the {\em total} combined data, and evaluate the desired function via a {\em boolean circuit} representation. When the number of parties and size of data is large, or when the functions to be computed are "lightweight" (e.g. touching only small portions of the data), these limitations completely obliterate feasibility of MPC as a solution.
In this talk, I will introduce a new class of techniques yielding MPC protocols whose parameters scale to the modern regime of massive data.
Lecture will be available to stream HERE.
Biography: Elette is currently a postdoctoral researcher at the Technion Israel Institute of Technology. Prior to the Technion, Elette received her Ph.D. from MIT under the guidance of Shafi Goldwasser, held a short-term postdoc at Cornell University, and completed her B.S. at Caltech in mathematics. Elette's research is in cryptography, focusing on methods of secure computation and distributed algorithm design.
Host: Computer Science Department
More Info: https://bluejeans.com/175107895
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
Event Link: https://bluejeans.com/175107895
-
CS Colloquium: Elias Bareinboim (UCLA) - Generalizability in Causal Inference
Thu, Mar 12, 2015 @ 09:45 AM - 10:50 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Elias Bareinboim, UCLA
Talk Title: Generalizability in Causal Inference
Series: CS Colloquium
Abstract: Empirical scientists seek not just surface descriptions of the observed data, but deeper explanations of why things happened the way they did, and how the world would be like had things happened differently. With the unprecedented accumulation of data (or, âbig dataâ), researchers are becoming increasingly aware of the fact that traditional statistical techniques, including those based on artificial intelligence and machine learning, must be enriched with two additional ingredients in order to construct such explanations:
1. the ability to integrate data from multiple, heterogeneous sources, and
2. the ability to distinguish causal from associational relationships.
In this talk, I will present a theory of causal generalization that provides a principled way for fusing pieces of empirical evidence coming from multiple, heterogeneous sources. I will first introduce a formal language capable of encoding the assumptions necessary to express each problem instance. I will then present conditions and algorithms for deciding whether a given problem instance admits a consistent estimate for the target effects and, if feasible, fuse information from various sources to synthesize such an estimate. These results subsume the analyses conducted in various fields in the empirical sciences, including âexternal validity,â âmeta-analysis,â âheterogeneity,â âquasi-experiments,â âtransportability,â and âsampling selection bias.â I will conclude by presenting new challenges and opportunities opened by this research.
The lecture will be available to stream Here.
Biography: Elias Bareinboim is a postdoctoral scholar (and was a Ph.D. student) in the Computer Science Department at the University of California, Los Angeles, working with Judea Pearl. His interests are in causal and counterfactual inferences and their applications. He is also broadly interested in artificial intelligence, machine learning, robotics, and philosophy of science. His doctoral thesis provides the first general framework for solving the generalizability problems in causal inference -- which has applications across all the empirical sciences. Bareinboim's recognitions include the Dan David Prize Scholarship, the Yahoo! Key Scientific Challenges Award, the Outstanding Paper Award at the 2014 Annual Conference of the American Association for Artificial Intelligence (AAAI), and the Edward K. Rice Outstanding Graduate Student.
Host: Computer Science Department
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
RASC Seminar Event: Prof. Carrick Detweiler (University of Nebraska-Lincoln) - Bringing Aerial Robots Closer to the Water: Sensing, Sampling, and Safety
Mon, Mar 23, 2015 @ 12:30 PM - 02:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Carrick Detweiler, University of Nebraska-Lincoln
Talk Title: Bringing Aerial Robots Closer to the Water: Sensing, Sampling, and Safety
Series: RASC Seminar Series
Abstract: Unmanned Aerial Vehicles (UAVs) are increasingly being used for everything from crop surveying to pipeline monitoring. They are significantly cheaper than the traditional manned airplane or helicopter approaches to obtaining aerial imagery and sensor data. The next generation of UAVs, however, will do more than simply observe. In this talk, I will discuss the challenges of using aerial robots very close to the water to obtain aerial water samples and sensor data from remote waters locations without needing to bring a boat to each location. When flying close to water, there is little time to react to errors and among obstacles. I will discuss automated software analysis techniques we are developing to detect and correct system errors to reduce risk and increase safety. I will focus on our recent work on the UAV-based water sampler system, but also discuss other applications we are pursuing, including using UAVs to recharge remotely deployed sensors and how we are using very low flying UAVs to monitor the growth of crops.
Biography: Dr. Carrick Detweiler is an Assistant Professor in the Computer Science and Engineering department at the University of Nebraska-Lincoln. He co-directs and co-founded the Nebraska Intelligent MoBile Unmanned Systems (NIMBUS) Lab at UNL, which focuses on developing software and systems for small aerial robots and sensor systems. Carrick obtained his B.A. in 2004 from Middlebury College and his Ph.D. in 2010 from MIT CSAIL. He is a Faculty Fellow at the Robert B. Daugherty Water for Food Institute at UNL and recently received the 2014 College of Engineering Henry Y. Kleinkauf Family Distinguished New Faculty Teaching Award. He is currently lead PI on NSF and USDA grants, including a National Robotics Initiative Grant. In addition to research activities, Carrick actively promotes the use of robotics in the arts through workshops and collaborations with the international dance companies Pilobolus and STREB.
Host: RASC
Location: Ronald Tutor Hall of Engineering (RTH) - 406
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Justin Solomon (Stanford University) - Transportation Techniques for Geometric Data Processing
Tue, Mar 24, 2015 @ 09:45 AM - 10:50 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Justin Solomon, Stanford University
Talk Title: Transportation Techniques for Geometric Data Processing
Series: CS Colloquium
Abstract: Modeling and understanding low- and high-dimensional data is a recurring theme in graphics, optimization, learning, and vision. Abstracting away application domains reveals common threads using geometric constructs like distances, similarities, and curvatures. This shared structure suggests the possibility of developing geometric data processing as a discipline in itself.
To this end, I will introduce optimal transportation (OT) as a versatile component of the geometric data processing toolkit. Originally proposed for minimizing the cost of shipping products from producers to consumers, OT links probability and geometry using distributions to encode geometric features and developing metric machinery to quantify their relationships.
To transition OT from theory to practice, I will show how to solve previously intractable OT problems efficiently on discretized domains and demonstrate a wide range of applications enabled by this new machinery. I will illustrate the advantages and challenges of OT for geometric data processing by outlining my recent work in geometry processing, computer graphics, and machine learning. In each case, I will consider optimization aspects of the OT problem for relevant geometric domains---including triangulated surfaces, graphs, and subsets of Euclidean space---and then show how the resulting machinery can be used to approach outstanding problems in surface correspondence, modeling, and semi-supervised learning.
This lecture will be streamed HERE.
Biography: Justin Solomon is a PhD candidate and teaching fellow in the Geometric Computing Group at Stanford University studying problems in shape analysis, machine learning, and graphics from a geometric perspective. His work is supported by the Hertz Foundation Fellowship, the NSF Graduate Research Fellowship, and the NDSEG Fellowship. Justin holds bachelors degrees in mathematics and computer science and an MS in computer science from Stanford. He has served as the lecturer for courses in graphics, differential geometry, and numerical methods; his forthcoming textbook entitled Numerical Algorithms focuses on applications of numerical methods across modern computer science. Before his graduate studies, Justin was a member of Pixar's Tools Research group. He is a pianist, cellist, and amateur musicologist with award-winning research on early recordings of the Elgar Cello Concerto.
Host: Computer Science Department
More Info: https://bluejeans.com/301312091/browser
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
Event Link: https://bluejeans.com/301312091/browser
-
Energy informatics distinguished seminar
Wed, Mar 25, 2015 @ 03:30 PM - 04:30 PM
Thomas Lord Department of Computer Science, Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
Speaker: Arvind, Massachusetts Institute of Technology
Talk Title: BlueDBM: A Multi-access, Distributed Flash Store for Big Data Analytics
Series: Energy Informatics Distinguished Seminar Series
Abstract: Complex analytics of the vast amount of data collected via social media, cell phones, ubiquitous smart sensors, and satellites is likely to be the biggest economic driver for the IT industry over the next decade. For many âBig Dataâ applications, the limiting factor in performance is often the transportation of large amount of data from hard disks to where it can be processed, i.e. DRAM. We will present BlueDBM, an architecture for a scalable distributed flash store which is designed to overcome this limitation in two ways. First, the architecture provides a high-performance, high-capacity, scalable random-access storage. It achieves high-throughput by sharing large numbers of flash chips across a low-latency, chip-to-chip backplane network managed by the flash controllers. Second, it permits some computation near the data via a FPGA-based programmable flash controller. We will present the preliminary results on accelerating complex queries using BlueDBM consisting of 20 nodes and up to 32 TB of flash.
Biography: Arvind is the Johnson Professor of Computer Science and Engineering at MIT. Arvindâs group, in collaboration with Motorola, built the Monsoon dataflow machines and its associated software in the late eighties. In 2000, Arvind started Sandburst which was sold to Broadcom in 2006. In 2003, Arvind co-founded Bluespec Inc., an EDA company to produce a set of tools for high-level synthesis. In 2001, Dr. R. S. Nikhil and Arvind published the book âImplicit parallel programming in pHâ. Arvind's current research focus is on enabling rapid development of embedded systems.
Arvind is a Fellow of IEEE and ACM, and a member of the National Academy of Engineering and the American Academy of Arts and Sciences.
Host: Viktor Prasanna
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 132
Audiences: Everyone Is Invited
Contact: Annie Yu
-
CS Colloquium: Suman Jana (Stanford) - Rise of the Planet of the Apps: Security and Privacy in the Age of Bad Code
Thu, Mar 26, 2015 @ 09:45 AM - 10:50 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Suman Jana, Stanford
Talk Title: Rise of the Planet of the Apps: Security and Privacy in the Age of Bad Code
Series: CS Colloquium
Abstract: Computing is undergoing a major shift. Third-party applications hosted in online software markets have become ubiquitous on all kinds of platforms: mobile phones, Web browsers, gaming devices, even household robots. These applications often include yet more third-party code for advertising, analytics, etc. These trends have dramatically increased the amount of bad code throughout the software stack - buggy code, malicious code, code that overcollects private information intentionally or by accident, overprivileged code vulnerable to abuse - as well as the amount of sensitive data processed by bad code.
In this talk, I will demonstrate that existing application platforms are ill-suited to dealing with bad code, thus causing security and privacy problems. I will then show how to isolate bad code without affecting its useful functionality, by redesigning the interfaces across the software stack and controlling the information released to the applications by the platform. I will also show how automated testing can identify bad code and help developers improve their applications.
The lecture will be streamed HERE.
Biography: Suman Jana is a postdoctoral researcher at Stanford University. He earned his PhD in 2014 from the University of Texas, where he was supported by the Google PhD Fellowship. He is broadly interested in identifying fundamental flaws in existing systems and building new systems with strong security and privacy guarantees. Suman received the 2014 PET Award for Outstanding Research in Privacy-Enhancing Technologies, Best Practical Paper Award from the 2014 IEEE Symposium on Security and Privacy (Oakland), Best Student Paper Award from the 2012 IEEE Symposium on Security and Privacy, and the 2012 Best Applied Security Paper Award.
Suman's research has been widely covered in popular media, and his code has been deployed at Google, Mozilla, and Apache.
Host: Computer Science Department
More Info: https://bluejeans.com/773593518
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
Event Link: https://bluejeans.com/773593518
-
RASC Seminar Event: Prof. Carrick Detweiler (University of Nebraska-Lincoln) - Bringing Aerial Robots Closer to the Water: Sensing, Sampling, and Safety
Thu, Mar 26, 2015 @ 12:30 PM - 01:30 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Carrick Detweiler, University of Nebraska-Lincoln
Talk Title: Bringing Aerial Robots Closer to the Water: Sensing, Sampling, and Safety
Series: RASC Seminar Series
Abstract: Unmanned Aerial Vehicles (UAVs) are increasingly being used for everything from crop surveying to pipeline monitoring. They are significantly cheaper than the traditional manned airplane or helicopter approaches to obtaining aerial imagery and sensor data. The next generation of UAVs, however, will do more than simply observe. In this talk, I will discuss the challenges of using aerial robots very close to the water to obtain aerial water samples and sensor data from remote waters locations without needing to bring a boat to each location. When flying close to water, there is little time to react to errors and among obstacles. I will discuss automated software analysis techniques we are developing to detect and correct system errors to reduce risk and increase safety. I will focus on our recent work on the UAV-based water sampler system, but also discuss other applications we are pursuing, including using UAVs to recharge remotely deployed sensors and how we are using very low flying UAVs to monitor the growth of crops.
Biography: Dr. Carrick Detweiler is an Assistant Professor in the Computer Science and Engineering department at the University of Nebraska-Lincoln. He co-directs and co-founded the Nebraska Intelligent MoBile Unmanned Systems (NIMBUS) Lab at UNL, which focuses on developing software and systems for small aerial robots and sensor systems. Carrick obtained his B.A. in 2004 from Middlebury College and his Ph.D. in 2010 from MIT CSAIL. He is a Faculty Fellow at the Robert B. Daugherty Water for Food Institute at UNL and recently received the 2014 College of Engineering Henry Y. Kleinkauf Family Distinguished New Faculty Teaching Award. He is currently lead PI on NSF and USDA grants, including a National Robotics Initiative Grant. In addition to research activities, Carrick actively promotes the use of robotics in the arts through workshops and collaborations with the international dance companies Pilobolus and STREB.
Host: RASC
Location: Ronald Tutor Hall of Engineering (RTH) - 406
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
-
CS Colloquium: Rong Ge (Microsoft Research) - Towards Provable and Practical Machine Learning
Tue, Mar 31, 2015 @ 09:45 AM - 10:50 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Rong Ge, Microsoft Research
Talk Title: Towards Provable and Practical Machine Learning
Series: CS Colloquium
Abstract: Many problems --- especially machine learning problems like sparse coding or topic modeling --- are hard in the worst-case, but nevertheless solved in practice by algorithms whose convergence properties are not understood. In this talk I will show how we can identify natural properties of "real-life" instances that allow us to design scalable algorithms for a host of well-known machine learning problems. Most of the talk will be focused on the sparse coding problem: a basic task in many fields including signal processing, neuroscience and machine learning where the goal is to learn a basis that enables a sparse representation of a given set of data, if one exists. Here we give a general framework for understanding alternating minimization which we leverage to analyze existing heuristics and to design new ones also with provable guarantees.
The lecture will be available to stream Here
Biography: Rong Ge obtained his Ph.D. at Princeton University, advised by Sanjeev Arora. Currently he is a post-doctoral researcher at Microsoft Research, New England. He is broadly interested in theoretical computer science and machine learning, especially applying algorithm design and analysis techniques to machine learning problems. The key thread running through his work is to identify natural properties of "real-life" instances that allow him to design scalable algorithms for several interesting machine learning problems including topic modeling and sparse coding.
Host: Computer Science Department
More Info: https://bluejeans.com/651721928
Location: Olin Hall of Engineering (OHE) - 132
Audiences: Everyone Is Invited
Contact: Assistant to CS chair
Event Link: https://bluejeans.com/651721928
