Dr. Daniel A. Jiménez
Associate Professor
Department of Computer Science and Engineering
Texas A&M University
4:10 p.m., Monday, January 14, 2013
Room 124, Bright Building
Abstract
Modern microprocessors achieve high performance through aggressive speculation. However, large amounts of energy and potential performance are lost by speculating fruitlessly. The two most important speculation techniques are caches and speculative execution.
Caches hold a subset of the blocks from the high-latency main memory, speculating that quick access to these blocks will benefit the program. Unfortunately, most blocks in the last-level cache will not be referenced again before they are removed from the cache. These dead blocks waste time and energy as they reduce the effective capacity of the cache.
Speculative execution mitigates pipeline control hazards by predicting the outcome of branches, allowing subsequent instructions to be fetched and executed down the predicted path. Many instructions will be wrongly executed before an incorrect prediction is discovered, again wasting time and energy.
This talk discusses novel techniques for reclaiming lost performance and energy through reducing speculation wasted by caches and speculative execution. The talk will also discuss ongoing projects and future research directions.
Biography
Daniel A. Jiménez is an Associate Professor in the Department of Computer Science and Engineering at Texas A&M University. Previously he was a full Professor and Chair of the Department of Computer Science at The University of Texas at San Antonio and Associate Professor (with tenure) in the Department of Computer Science at Rutgers University. His research focuses on microarchitecture and low-level compiler optimizations. He introduced and developed the perceptron branch predictor which has inspired the design of two implemented microarchitectures: the AMD "Bobcat" core and the Oracle SPARC T4.
Daniel earned his B.S. (1992) and M.S. (1994) in Computer Science at The University of Texas at San Antonio and his Ph.D. (2002) in Computer Sciences at The University of Texas at Austin. From 2002 through 2007, Daniel was an Assistant Professor in the Department of Computer Science at Rutgers. In 2005 Daniel took sabbatical leave at the Technical University of Catalonia (UPC) in Barcelona, Catalonia, Spain. In 2008 he was promoted to Associate Professor with tenure at Rutgers. He returned to his native Texas to take a position at UT San Antonio. He recently returned from a second sabbatical leave in Spain at the Barcelona Supercomputing Center and was General Chair of the 2011 IEEE HPCA conference. He consults with Samsung on very cool projects that Samsung won't let him talk about. He is an NSF CAREER award recipient and ACM Distinguished Scientist.
Faculty Contact: Dr. Lawrence Rauchwerger (rwerger [at] cse.tamu.edu)
Dr. Quentin F. Stout
Professor
Computer Science and Engineering
University of Michigan
4:10 p.m., Wednesday, January 16, 2013
Room 124, Bright Building
Abstract
In the 50's von Neumann developed the cellular automata model of parallel computing, partially motivated by models of crystal growth. In the 70's and 80's extensions of cellular automata were used to model the computational and physical constraints of VLSI. In the 10's they are relevant yet again as models of computation and locality.
Massively parallel systems will be constrained by power consumption, where "flops are free" but data movement isn't, and much of the silicon may need to be dark. Fortunately new capabilities are emerging, such as 3-d modules. The talk will discuss the evolving models and show some asymptotic limits of computing in 3-space vs. 2-space and tradeoffs between time, space, and power.
Biography
Quentin Stout is a professor of Computer Science and Engineering at the University of Michigan, where he has been since 1984. Previously he was a professor of Mathematical Sciences at State University of New York at Binghamton. His PhD from Indiana University, and BA from Centre College, are both in mathematics. Prior to that he was a nerd in the public schools of Euclid, Ohio.
His primary research interests are in parallel computing, both theory and applications. At Michigan he is a co-director of the Center for Space Environment Modeling and a co-PI of the Center for Radiative Shock Hydrodynamics. He is also on the executive committee of the Earth Systems Modeling Framework, which is incorporated into the operational models of the National Weather Service.
Faculty Contact: Dr. Lawrence Rauchwerger (rwerger [at] cse.tamu.edu)
Dr. Bruce W. Porter
Professor and Chairman
Department of Computer Sciences
University of Texas at Austin
4:10 p.m., Wednesday, January 23, 2013
Room 124, Bright Building
Abstract
Knowledge-based systems are now capable of advanced question-answering in broad fields, such as chemistry and biology. They use automated reasoning and explanation-generation techniques to answer hard, novel questions with coherent explanations expressed in English. This talk will describe Project Halo, which started with a competition to build a system capable of answering questions like those on the AP Chemistry exam, and then progressed to building knowledge-capture tools to enable domain experts, with no training in computer science, to build such systems independently.
Biography
Bruce Porter is a Professor in the Department of Computer Science at the University of Texas at Austin, and he is currently the Department Chairman. His teaching and research spans the breadth of knowledge-based question answering, including representation, reasoning, learning and explanation generation. Porter's research group has played major roles in DARPA's projects on Rapid Knowledge Formation and Learning by Reading, as well as Vulcan's Project Halo. Dr. Porter has served as the chairman of numerous conferences including the National Conference on Artificial Intelligence (AAAI 2001) and the Innovative Applications of Artificial Intelligence conference (IAAI 2006). He earned a Presidential Young Investigator Award from the National Science Foundation in 1988.
Faculty Contact: Dr. Duncan M. (Hank) Walker (walker [at] cse.tamu.edu)
Dr. Ünal "Zak" Sakoglu
Assistant Professor
Computer Science Department & Information Systems
Texas A&M University - Commerce
4:10 p.m., Monday, January 28, 2013
Room 124, Bright Building
Abstract
Functional Magnetic Resonance Imaging (fMRI), which involves taking snapshots of brain every few seconds while the subject is doing certain tasks in the scanner, has been a very useful and noninvasive imaging modality to study the brain's functioning. Along with univariate methods, application of multivariate data analysis and pattern classification algorithms have lead to better detection of brain activity and better understanding of the memory function from fMRI data. In this talk, I will present an fMRI memory experiment and how multivariate classification methods can be used along with univariate methods to better detect the brain activity and can be used to do a successful classification of whether the subject were looking at pictures or words.
Biography
Ünal "Zak" Sakoglu is currently an Assistant Professor at Texas A&M University - Commerce Computer Science Department. He had his BS in Electrical-Electronics Engineering from Bilkent University, Ankara, Turkey, and MS & PhD degrees in Electrical and Computer Engineering from University of New Mexico in Albuquerque, NM. His graduate research involved developing signal and image processing algorithms for better multispectral classification with infrared array sensors. He did his post-doctoral training at University of New Mexico Neurology Department BRAIN Imaging Center and Mind Research Network in Albuquerque, where he developed and applied data analysis & classification techniques to functional magnetic resonance imaging data.
He worked as Research Scientist at UT Southwestern Medical Center Neuroradiology Department, at Abbott Laboratories Translational Imaging Group, and UT Dallas Center for Vital Longevity, where he analyzed different modalities of medical imaging data such as PET/CT, SPECT/CT, MRI and fMRI, during these positions. He is currently working on development and application of multivariate pattern classification, data-mining and machine-learning methods to fMRI data in order to understand the way human brain is functioning and how it is effected by different brain conditions (different stimuli, age, disease, etc.). He is also working on developing fMRI signal simulation and better visualization techniques for better classification of multidimensional data.
Faculty Contact: Dr. Yoonsuck Choe (choe [at] cse.tamu.edu)
Dr. José Daniel García Sánchez
Associate Professor
Computer Architecture and Technology Area
Universidad Carlos III de Madrid
4:10 p.m., Wednesday, February 6, 2013
Room 124, Bright Building
Abstract
A Parallel File System is a mechanism that allows multiple processes in an application to access data of a shared file in parallel, delivering high performance I/O bandwidth. Typically a Parallel File System spreads files data among multiple disks located in different nodes of a network or cluster. Parallel file systems allow the usage of standardized APIs (as POSIX) or specialized APIs (MPI-IO), and provide parallel access to files, increase the I/O bandwidth, load balancing among I/O nodes and disks, and higher storage capacities.
In this talk I will present Expand, a Parallel File System developed at UC3M. Unlike other solutions, Expand provides a high performance storage system by using standard protocols and servers, allowing easy integration of heterogeneous systems, reuse and aggregation of existing resources while providing parallel data access. Key design elements in Expand will be presented as well as its adaptations for cluster, grid and volunteer computing.
Biography
Prof J. Daniel Garcia received his Engineering Degree in Computer Science from Madrid Technical University and his PhD in Computer Science and Engineering from University Carlos III of Madrid. Before completing his PhD he worked for the industry in projects for major companies in Spain and Germany including Telefonica, British Telecom, ING Bank, SIEMENS and DMR Consulting. He has also been Lecturer at Pontifical University of Salamanca. Since 2006 he has been an Associate Professor in the Computer Science and Engineering Department at University Carlos III of Madrid, where he is a member of the Computer Architecture, Communications and Systems Group (http://www.arcos.inf.uc3m.es). He has published more than 50 journal and conference papers and has edited several special issues for the Journal of Supercomputing. He has participated in many technology transfer contracts and publicly funded research projects, before and after joining academia. He has been visiting researcher at University of Modena, Italy. He was General Chair of IEEE ISPA 2012 and ICA3PP 2010 and has served in PC of many other conferences related to parallel and distributed systems. Since 2008 he has been the Spanish Head of delegation in the ISO C++ Standards Committee. He is a Senior Member of the IEEE and a member of the ACM.
Faculty Contact: Dr. Bjarne Stroustrup (bs [at] cse.tamu.edu)
Dr. Jian Li
Research Staff Member, IBM Austin Research Laboratory
Adjunct Assistant Professor
Department of Computer Science and Engineering
Texas A&M University
4:10 p.m., Monday, February 11, 2013
Room 124, Bright Building
Abstract
Data continue a massive expansion in scale, diversity, and complexity. Data underpin activities in all sectors of society. As informatics is the science of information, analytics is the science of data analysis. Achieving the full transformative potential from the use of data in this increasingly digital world requires not only new data analysis algorithms but also a new generation of systems and distributed computing environments to handle the dramatic growth in the volume of data, the lack of structure for much of it and the increasing computational needs of massive-scale analytics. In this talk, I will share my view on Big Data, present an update on IBM's Big Data platform, and discuss a few research topics in the Big Data era.
Biography
Jian Li is a research staff member at IBM Research in Austin and the Chief Architect of Big Data Systems (China) when he is on company assignment with IBM Greater China Group. He is in transition to an assignment in IBM Growth Market Unit Big Data Center of Competency. He holds a Ph.D. degree in Electrical and Computer Engineering from Cornell University. His current research centers on Big Data Analytics systems and their applications. He has worked in the areas of architectural support for power- and variation-aware computing, interconnection network design for high-performance computing systems, workload-driven three-dimensional integration (3DI) architecture, cost-driven 3DI, architectural applications of non-volatile memory (NVM) and storage class memory (SCM), energy-efficient interconnection networks, data center networks, workload-optimized systems and Big Data analytics, with a strong emphasis on entertaining his family members in Austin, Texas. He holds an adjunct position at CSCE, Texas A&M University.
Faculty Contact: Dr. Lawrence Rauchwerger (rwerger [at] cse.tamu.edu)
Dr. Sujatha Kashyap
Performance Architect
IBM
4:10 p.m., Monday, February 18, 2013
Room 124, Bright Building
Abstract
Several forces have dramatically changed the definition of computing performance over the past five years. Compute capacity has coalesced into large data centers, allowing the application of formidable consolidated capacity to meet an enterprise's IT requirements. The global economic recession slashed IT budgets, causing a shift in interest from absolute performance to using the fewest possible resources to meet throughput and quality of service requirements. The rapidly growing field of structured and unstructured analytics is continuously creating a slew of new products and deployment patterns, yielding new targets for performance analysis and optimizations. In this talk, we will cover some of the current sources of inefficiencies in these environments, which provide opportunities for improved system design.
Biography
Sujatha Kashyap is a performance architect for the POWER family of systems from IBM. She has worked on understanding and improving the performance of workloads on various POWER architectures for over a decade. Her experience spans online transaction processing systems, highly virtualized environments, edge-of-network appliances, Big Data frameworks, business analytics applications, and several others. She is currently focused on understanding emerging workload areas including in-memory databases and various analytics platforms, and their implications on the design of future POWER systems. Sujatha received her M.S. in Computer Science from Texas A&M University and her Ph.D. in Computer Engineering from the University of Texas at Austin.
Faculty Contact: Dr. Lawrence Rauchwerger (rwerger [at] cse.tamu.edu)
Dr. Frank Vahid
Professor
Dept. of Computer Science & Engineering
University of California, Riverside
4:10 p.m., Wednesday, February 20, 2013
Room 124, Bright Building
Abstract
Designing computer-based medical devices like pacemakers or ventilators is hard, in part because testing can't be done on real humans. PC-based simulations are slow and inaccurate. Using physical mockups, like connecting a ventilator device to a balloon acting as a lung, can't support sufficiently diverse scenarios, like fluid in the lungs. This talk describes joint UCR/UCI work on developing "digital mockups" — models of physiological systems that execute in real-time, supporting thorough testing of device software. Digital mockups are also useful in building complete human simulators, used today medical and nursing schools. Digital mockups thus represent an important component in a science of design for cyber-physical systems. We show that FPGAs (field-programmable gate arrays) — widely-available programmable chips having a unique execution approach that we'll describe — are an excellent match for executing physiological models. The talk describes the synthesis approach we've evolved over the past years to automatically compile models, consisting of thousands of differential equations, into massively-parallel networks of processing elements on FPGAs. The approach yields the fastest-known execution of physical models on mainstream computing devices, with the best performance/cost ratio too.
The talk will conclude with a brief summary of other efforts from our lab, including automated assistive monitoring for live-alone senior citizens, the use of technology to combat drunk driving, and new web-based animated interactive learning materials.
Biography
Frank Vahid is a Professor of Computer Science and Engineering at the University of California, Riverside (B.S. 1988 Univ. of Illinois in 1988, M.S./Ph.D. 1990/1994 Univ. of California, Irvine. He is author of several textbooks on embedded systems and digital design. His current research interests include creating technologies for cyber-physical systems (http://www.cs.ucr.edu/~vahid/digitalmockups), developing customizable assistive monitoring systems for home-alone aging/disabled persons and their caretakers (http://www.cs.ucr.edu/~vahid/assistivemonitoring), combating drunk driving (http://duicam.org), and creating the next generation of web-based interactive animated learning material (http://zyante.com).
Faculty Contact: Dr. Rabi Mahapatra (rabi [at] cse.tamu.edu)
Dr. Michael Gschwind
STSM & Senior Manager, System Architecture
IBM STG & IBM TJ Watson Research Center
4:10 p.m., Monday, February 25, 2013
Room 124, Bright Building
Abstract
The Blue Gene/Q system represents the third generation of optimized high-performance computing Blue Gene solution servers and provides a platform for continued growth in HPC performance and capability. Blue Gene/Q started with a new design of the hardware platform, while retaining and significantly expanding an established, trusted and successful software environment.
To deliver a system that enables users to fully exploit the promise of high-performance computing for both traditional HPC applications and new commercial application areas, the Blue Gene/Q system architecture combines hardware and software innovations to overcome traditional bottlenecks, most famously the memory and power walls which have become emblematic of modern computing systems. At the same time, to deliver a platform for sustainable petascale computing, and beyond to exascale, we had to address a new set of “walls” with the many innovations described below: a scalability wall, a communication wall, and a reliability wall.
The new Blue Gene/Q system increases overall system performance with a new node architecture: Each node offers more thread-level-parallelism with a coherent SMP node consisting of eighteen 64-bit PowerPC cores with 4-way simultaneous multithreading. Each core provides for better exploitation of data-level parallelism with a new 4-way quad-vector processing unit (QPU). The memory subsystem integrates memory speculation support which can be used to implement both Transactional Memory and Speculative Execution programming models.
The compute nodes are connected in a five dimensional torus configuration using 10 point-to-point links, and a total network bandwidth of 44 GB/s per node. The on-chip messaging unit provides an optimized interface between the network routing logic and the memory subsystem, with enough bandwidth to keep all the links busy. It also offloads communication protocol processing by implementing collective broadcast and reduction operations, including integer and floating point sum, min and max. Built on the Blue Gene hardware design is an efficient software stack that builds on several generations of Blue Gene software interfaces, while extending these capabilities and adding new functions to support new hardware capabilities. The hardware functions were designed with a focus on providing efficient primitives upon which to build the rich software environment.
To ensure reliable operation of a petascale system, reliability has to be a pervasive design consideration. At the architecture level, new QPX store-and-indicate instructions support the detection of programming errors. To ensure reliable operation in the presence of transient faults, we conducted exhaustive single event upset simulations based on fault injection into the simulated design. The operating system was structured to use firmware in a small on-chip boot eDRAM to avoid silent system hangs.
Together, the hardware and software innovations pioneered in Blue Gene/Q give application developers a platform and framework to develop and deploy sustained petascale computing applications. These petascale applications will allow its users to make new scientific discoveries and gain new business insights, which will be the true measure of the success of the new Blue Gene/Q systems.
Biography
Dr. Michael Gschwind is a Senior Technical Staff Member and Senior Manager of System Architecture in the IBM Systems and Technology Group. In his dual role as a technical leader and manager, he is responsible for leading the architecture evolution of IBM's mainframe System z and Power systems and manages the architecture teams for both System z and Power brands. Previously, Dr. Gschwind served as Blue Gene Floating Point Chief Architect and Lead and as lead for Blue Gene/Q core soft error resilience.
Dr. Gschwind served as IFU and IDU lead and chief microarchitect for the Komal core which is the foundation for Power7 systems and as architecture lead for Productive, Easy-to-use, Reliable Computing Systems (PERCS) for DARPA's High Productivity Computing Systems (HPCS) initiative where he served as lead architect for IBM’s vector/scalar SIMD architecture (VSX). Dr. Gschwind helped initiate the Cell project, served as one of its lead architects for the Cell definition, developed the first Cell compiler and served as technical lead for the software team. He also helped create the media extensions for the Xbox360 chip. Dr. Gschwind received his PhD from Technische Universität Wien. Dr. Gschwind is an IEEE Fellow, member of the ACM SIGMICRO executive committee, an IBM Master Inventor and a member of the IBM Academy of Technology and was named as an industry-leading “IT Innovator and Influencer” by InformationWeek in 2006.
In addition to serving as a member of the ACM SIGMICRO board, Dr. Gschwind has served as Program Co-Chair of ACM Computing Frontiers 2008, as General Chair of the ACM International Conference on Supercomputing 2009, and as Program Co-Chair of the ACM/IEEE/IFIP International Conference on Parallel Architectures and Compilation Techniques 2010. Dr. Gschwind has also served on the Steering Committees of the ACM Computing Frontiers, ACM International Conference on Supercomputing and ACM/IEEE/IFIP International Conference on Parallel Architectures and Compilation Techniques conference series.
Faculty Contact: Dr. Lawrence Rauchwerger (rwerger [at] cse.tamu.edu)
Dr. Joe D. Warren
Chair and Professor of Computer Science
Rice University
4:00 p.m., Tuesday, February 26, 2013
Room 124, Bright Building
Abstract
Interest in massive open online courses (MOOCs) has exploded in the last 18 months. Educational consortiums such as Coursera and edX now offer free classes from top-ranked universities such as Stanford and MIT that are open to students of all backgrounds.
In late fall of 2012, the speaker along with three other faculty members in the Department of Computer Science at Rice University taught Rice's first MOOC on Coursera. In this talk, the speaker will share his experience in preparing and teaching this class. He will also offer his perspective on the effect of open online education on both students and faculty as well as its potential impact on universities.
Biography
Professor Warren attended Rice University and received an undergraduate degree in Computer Science in 1983. He then attend graduate school at Cornell University and received his PhD in Computer Science in 1986. After graduation, he returned to Rice University where he has been a Professor of Computer Science ever since. Professor Warren has served as Chair of the Department of Computer Science since 2008.
Professor Warren's research interests include geometric modeling and design, computational geometry, computer graphics and educational technology. His current research focuses on discrete multi-resolution methods for modeling smooth shape and their applications to interesting applied problems. Together with Henrik Weimer, Dr. Warren is the author of the book Subdivision Methods for Geometric Design.
Faculty Contact: Dr. Duncan M. (Hank) Walker (walker [at] cse.tamu.edu)
Sean Parent
Principal Scientist and Software Architect
Adobe Mobile Digital Imaging Group
4:10 p.m., Wednesday, March 6, 2013
Room 124, Bright Building
Abstract
This talk explains why (and how) to implement polymorphism without inheritance in C++. The talk contains many C++ tips and techniques, including many new features from C++11. During the course of that talk a key feature from Photoshop will be demonstrated and implemented.
Biography
Sean Parent is a principal scientist and software architect for mobile digital imaging at Adobe. Sean has been at Adobe since 1993 when he joined as a senior engineer working on Photoshop and later managed Adobe’s Software Technology Lab. In 2009 Sean spent a year at Google working on Chrome OS before returning to Adobe. From 1988 through 1993 Sean worked at Apple, where he was part of the system software team that developed the technologies allowing Apple’s successful transition to PowerPC.
Faculty Contact: Dr. Bjarne Stroustrup (bs [at] cse.tamu.edu)
Xiaohu Guo
Associate Professor
Department of Computer Science
University of Texas at Dallas
4:10 p.m., Monday, March 18, 2013
Room 124, Bright Building
Abstract
Physics-based simulation is one of the most important approaches to achieve "realism" for many real-time computer graphics applications such as computer games. In this talk, I will introduce the basic theory behind physics-based elastic deformation, its computational approaches, and the main challenges. Then I will introduce our latest work of an efficient, flexible, and robust framework for simulating 3D elastic deformation of soft objects based on domain decomposition and subspace model reduction. The soft objects are decomposed into multiple domains, and the dynamic ordinary differential equations are linearized domain-wise and projected into a carefully-designed local subspace. The domain coupling is fully satisfied in these subspace without any artifact. The comparison with the other state-of-the-art approaches on multi-domain subspace deformation will be further discussed.
Biography
Xiaohu Guo is an associate professor of computer science at the University of Texas at Dallas. He received the PhD degree in computer science from the State University of New York at Stony Brook in 2006. His research interests include computer graphics, animation and visualization, with an emphasis on geometric, and physics-based modeling. His current researches at UT-Dallas include: spectral geometric analysis, deformable models, centroidal Voronoi tessellation, mesh generation, medical image analysis, etc. He received the NSF CAREER Award in 2012.
Faculty Contact: Dr. John Keyser (keyser [at] cse.tamu.edu)
Bruno Siciliano
Professor of Automatic Control
Department of Electrical Engineering and Information Technology
University of Naples Federico II
4:10 p.m., Monday, March 25, 2013
Room 124, Bright Building
Abstract
Robots! Robots on Mars and in oceans, in hospitals and homes, in factories and schools; robots fighting fires, making goods and products, saving time and lives. Robots today are making a considerable impact on many aspects of modern life, from industrial manufacturing to healthcare, transportation, and exploration of the deep space and sea. Tomorrow, robots will be as pervasive and personal as today’s personal computers. The dream to create machines that are skilled and intelligent has been part of humanity from the beginning of time. This dream is now becoming part of our world’s striking reality. Beyond its impact on physical robots, the body of knowledge robotics has produced is revealing a much wider range of applications reaching across diverse research areas and scientific disciplines, such as: biomechanics, haptics, neurosciences, virtual simulation, animation, surgery, and sensor networks among others. In return, the challenges of the new emerging areas are proving an abundant source of stimulation and insights for the field of robotics. It is indeed at the intersection of disciplines that the most striking advances happen. Today, new communities of users and developers are forming, with growing connections to the core of robotics research. A strategic goal for the robotics community is one of outreach and scientific cooperation with these communities. Future developments and expected growth of the field will largely depend on the research community’s abilities to achieve this objective. This talk revisits 50 years and more of research and development in robotics and provides the active trends and perspectives of the field.
Biography
Bruno Siciliano is Professor of Control and Robotics, and Director of the PRISMA Lab in the Department of Electrical Engineering and Information Technology at University of Naples Federico II. His research interests include force and visual control, human-robot interaction and service robotics. He has co-authored 7 books, 70 journal papers, 170 conference papers and book chapters. He has delivered 100 invited lectures and seminars at institutions worldwide, and he has been the recipient of several awards. He is a Fellow of IEEE, ASME and IFAC. He has served on the editorial boards of several peer-reviewed journals and has been chair of program and organizing committees of several international conferences. He is Co-Editor of the Springer Tracts in Advanced Robotics, and of the Springer Handbook of Robotics, which received the PROSE Award for Excellence in Physical Sciences & Mathematics and was also the winner in the category Engineering & Technology. His group has been granted twelve European projects. Professor Siciliano is the Past-President of the IEEE Robotics and Automation Society.
Faculty Contacts: Dr. Richard A. Volz (volz [at] cse.tamu.edu) and Dr. Nancy Amato (amato [at] cse.tamu.edu)
CSCE 681 Graduate Seminar:
*Lecture made possible by the Center for Robot–Assisted Search and Rescue at Texas A&M University.
Rick Lynch
Lieutenant General, US Army (Retired)
Executive Director, University of Texas at Arlington Research Institute (UTARI)
4:10 p.m., Wednesday, March 27, 2013
Room 124, Bright Building
Abstract
Rick Lynch, Lieutenant General, US Army (Retired) and Executive Director at the University of Texas at Arlington Research Institute (UTARI) will lecture about using robotic technology to help people with disabilities (including our Wounded Warriors) and to help human beings in the performance of dirty, dull and dangerous tasks (military applications, agriculture, manufacturing, etc).
Biography
Lieutenant General Rick Lynch graduated from the United States Military Academy in 1977 and was commissioned as a Regular Army Engineer Officer. As an Engineer, he commanded both a Combat Engineer Company and a Mobile Assault Bridge Company in the 17th Engineer Battalion, 2nd Armored Division. Lieutenant General Lynch later branch transferred to Armor. As an Armor field grade officer he was assigned to the 11th Armored Cavalry Regiment where he served as an S3 (Operations Officer) for the 1st Squadron and later as the Regimental Adjutant and the Regimental Executive Officer.
After retirement, Lieutenant General Lynch continued his service to others in a variety of capacities. He just finished a book about adaptive leadership under difficult circumstances (Adapt or Die, the Leadership Principles of an American General) that will be published in the fall of 2013. He makes presentations around the Nation on leadership, leader development, diversity and change management. He is also the Executive Director of the University of Texas at Arlington Research Institute (UTARI). UTARI focuses on the pursuit of advanced technology to help humanity, in the areas of robotics, biomedical research, advanced manufacturing, and energy/water/environment.
More information can be found at http://www.uta.edu/utari/people/Lynch%20Rick.html.
Faculty Contact: Dr. Robin Murphy (murphy [at] cse.tamu.edu)
Dr. Bjarne Stroustrup
University Distinguished Professor and
College of Engineering Endowed Chair in Computer Science
Texas A&M University
4:10 p.m., Monday, April 1, 2013
Room 124, Bright Building
Abstract
C++ allows you to write better code faster. By "better" I mean maintainable code with fewer errors than was possible in C++98. C++11 allows you to write less code for a given problem and have it run faster. By "faster" I mean getting real-world code to run as fast as or faster than hand-tuned C, as fast as or faster than code written in any modern language I know of, sometimes much faster. This can be done today, using currently shipping compilers.
But most people are stuck in a 1970s or 1980s mindset, can we catch up to C++11? Worse, many people are stuck in a mess of "legacy code" creating a framework of constraints that discourage the use of 21st century facilities.
My aim in this talk is not to enumerate the C++11 features or to go into great technical detail on a select feature. My aim is to show how the best practices for C++ design and programming is better supported by C++11 than by earlier versions. To do that, I discuss small code examples. I expect to use the concurrency library, standard containers, and chrono. I expect to use initializer lists, move semantics, variadic templates, lambda expressions, and type aliases. As usual, RAII (Resource Acquisition Is Initialization) will feature large.
Biography
Bjarne Stroustrup is the designer and original implementer of C++ and the author of several books (incl. "Programming — Principles and Practice using C++" and "The C++ Programming Language) and many popular and academic publications. His research interests include distributed systems, design, programming techniques, software development tools, and programming languages. He is actively involved in the ISO standardization of C++. Dr. Stroustrup is a University Distinguished Professor at Texas A&M University and the holder of the College of Engineering Chair in Computer Science. He retains a link with AT&T Labs-Research as an AT&T Fellow. He is a member of the US National Academy of Engineering, and IEEE Fellow and an ACM fellow. Born in Aarhus Denmark 1950. Cand.Scient. (Mathematics and Computer Science), 1975, University of Aarhus Denmark. Ph.D. (Computer Science) 1979, Cambridge University, England. www.research.att.com/~bs.
Faculty Contact: Dr. Lawrence Rauchwerger (rwerger [at] cse.tamu.edu)
This lecture is jointly sponsored by the Institute for Applied Mathematics and Computational Science (IAMCS).
Dr. Stephanie Forrest
Professor of Computer Science
University of New Mexico
4:10 p.m., Monday, April 15, 2013
Room 124, Bright Building
Abstract
Computer programmers like to think of software as the product of intelligent design, carefully crafted to meet well-specified goals. In reality, large software systems evolve inadvertently through the actions of many individual programmers, often leading to unanticipated consequences. Because software is subject to constraints similar to those faced by evolving biological systems, we have much to gain by viewing software through the lens of evolutionary biology. The talk will highlight recent research applying the mechanisms of evolution quite directly to the problem of repairing software bugs, including security vulnerabilities. It will describe an automated method for repairing errors in off-the-shelf, legacy programs without formal specifications, program annotations, or special coding practices.
Biography
Stephanie Forrest is Regents Professor of Computer Science at the University of New Mexico in Albuquerque, and a member of the Santa Fe Institute External Faculty. Her research studies adaptive systems, including immunology, evolutionary computation, biological modeling, computer security, and software. Professor Forrest received M.S. and Ph.D. degrees in Computer and Communication Sciences from the University of Michigan and a B.A. from St. John's College. Before joining UNM in 1990 she worked for Teknowledge Inc. and was a Director's Fellow at the Center for Nonlinear Studies, Los Alamos National Laboratory.
Faculty Contact: Dr. Nancy Amato (amato [at] cse.tamu.edu)
This lecture is jointly sponsored by the Center for Emergency Informatics Distinguished Lecture program.
John Crowley
Research Fellow
Harvard Humanitarian Initiative
Cambridge, Massachusetts
4:10 p.m., Wednesday, April 17, 2013
Room 124, Bright Building
Abstract
With minimal funding or support, mavericks are creating an ecosystem of open data, some of it collected via participatory means (crowdsourcing) and some of it unfrozen from archives of large institutions. And they are intertwingling these information resources in ways that challenge traditional notions of authoritative knowledge, privacy, mutual aid, and the role of disaster survivors. This lecture will map the mavericks and explore how the emerging information ecosystem may drive change in the way that responders make decisions. It will include an exploration of how government agencies are using collective intelligences in mapping (US State Department and OpenStreetMap) and data munging (USAID/Standby Task Force), as well as the ways in which these mashups are orchestrated (Joint Interagency Field Explorations at Camp Roberts). It will also explore emerging efforts at the UN, including the Humanitarian Exchange Language.
Biography
John Crowley is a research fellow at the Harvard Humanitarian Initiative and (in a former life) both a professional cellist and historian. He works as a synergist, connecting formal humanitarian institutions with informal communities and grassroots efforts. Using principles he learned as a musician and humanist, he builds interfaces between unlikely partners: civilian and military, high and low tech, commercial and not-for-profit.
John is an alumnus of the Kennedy School of Government's Mid-Career MPA program, where he was the Robert C. Seamans, Jr. Fellow in Science, Technology, and Public Policy. He holds an MA in interdisciplinary humanities from the University Professors at Boston University, and a MusB in Cello Performance and Music History & Literature from the Boston University School of Music. He tweets at @jcrowley.
Faculty Contact: Dr. Robin Murphy (murphy [at] cse.tamu.edu)
Dr. Frank Shipman
Professor
Department of Computer Science and Engineering
Texas A&M University
4:10 p.m., Monday, April 22, 2013
Room 124, Bright Building
Abstract
Today, more content is authored in social media systems than in traditional media. By definition, much of this information is collaboratively authored such that the individual contributions are deeply intertwined, thus complicating information property rights. Who can save, share, repurpose, or edit/delete this content? Current policies and systems are often developed based on an understanding of information ownership in traditional media. We surveyed more than 2000 social media users to explore their attitudes about content ownership and their practices concerning social media content. The results show how technology, the authoring context, the potential action concerning the content, and user's understanding of policy affect use. These results point towards the need for integrated design of social media technology and policy.
Biography
Frank is Associate Director of the TEES Center for the Study of Digital Libraries and a Professor in the Department of Computer Science and Engineering at Texas A&M University. He has worked at the intersection of human-computer interaction, artificial intelligence, information retrieval, and multimedia for 25 years.
Faculty Contact: Dr. Nancy Amato (amato [at] cse.tamu.edu)
This lecture is jointly sponsored by the Center for Emergency Informatics Distinguished Lecture program.
Dr. Dana Yoerger
Senior Scientist
Applied Ocean Physics & Engineering
Woods Hole Oceanographic Institution
4:10 p.m., Wednesday, April 24, 2013
Room 124, Bright Building
Abstract
Since Heezen and Tharp published the first comprehensive map of the entire ocean floor in 1977, our understanding of the deep seafloor has progressed rapidly. As in any scientific endeavor, enabling technologies have played critical roles. At the time, the prospect of publishing peer-reviewed work based on observations from submersibles was greeted in many circles with skepticism. Further innovations, including tow sleds, remotely operated vehicles, and autonomous platforms were likewise initially dismissed by the mainstream oceanographic community as impractical, too expensive, or ill-suited to examining important scientific questions. Today, the partnership between advanced technology and scientific discovery is unquestioned.
In this talk, I will review some of that history, especially aspects I have observed and participated in since I entered the field in 1979. These include the introduction of dynamic positioning on our research vessels and the introduction of Remotely Operated Vehicles (ROVs) in the 1980s. In that period, we improved our understanding of underwater vehicle dynamics and introduced accurate localization, automatic control, and precise mapping into the ROVs. These lessons would later prove critical to our early autonomous vehicles. Our autonomous vehicle work with with the AUV ABE began focused on repeated long term observations of vent sites, but was realigned to Mid-Ocean Ridge magnetic and bathymetric survey in response to funded opportunities and the willingness of our scientific collaborators to take some risks. We later expanded that work to include hydrothermal vent discovery and survey. Today, our AUVs can perform a wide range of tasks, including near-bottom photography, in-situ chemical survey, and can map bathymetry and plume structures with high resolution. They can make rudimentary decisions on their own. We have demonstrated these capabilities outside our initial Mid-Ocean Ridge focus to include tasks like the quantification of the deep hydrocarbon plumes arising from the BP oil spill and the search for sensitive deep sea coral sites that may have been impacted by the spill.
The future of seafloor exploration and the dual problem of under ice survey will take elements from both ROVs and AUVs. AUVs operating with full autonomy will no doubt be applied to problems where those attributes are needed. Depending on the task and setting, untethered vehicles can also be operated under direct supervision of human operators, much like ROVs. These developments are enabled by dramatic improvements in through water communications, including lightweight optical fiber tethers, improved acoustic communications, and through-water optical communications. The talk will conclude with speculation of how these platforms will evolve and how they will enable new areas of scientific inquiry.
Biography
Dr. Dana Yoerger is a Senior Scientist at the Woods Hole Oceanographic Institution and a researcher in robotics and unmanned vehicles. He supervises the research and academic program of graduate students studying oceanographic engineering through the MIT/WHOI Joint Program in the areas of control, robotics, and design. Dr. Yoerger has been a key contributor to the remotely-operated vehicle JASON; to the Autonomous Benthic Explorer known as ABE; most recently, to the autonomous underwater vehicle, SENTRY; and the hybrid remotely operated vehicle, NEREUS which reached the bottom of the Mariana Trench in 2009. Dr. Yoerger has gone to sea on over 70 oceanographic expeditions exploring the Mid-Ocean Ridge, mapping underwater seamounts and volcanoes, surveying ancient and modern shipwrecks, and studying the environmental effects of the Deepwater Horizon oil spill. He was the 2009 recipient of the Lockheed Award for Ocean Science and Engineering and serves on the Research Board for BP's Gulf of Mexico Research Initiative.
Faculty Contact: Dr. Robin Murphy (murphy [at] cse.tamu.edu)
CSCE 681 Graduate Seminar (only required for new graduate students):
4:10-6:00 p.m., Monday August 27, 2012
Room 124, Bright Building
Abstract
This meeting will concentrate on the essentials the students will need to settle in. It will include an introduction to departmental administration (staff who's who, payroll, mailboxes, phones), the computing resources (computer use/accounts, printer quotas, lab access/tours), the academic advising staff and resources, the TAMU honor system, TAMU Libraries, Graduate Teaching Academy, Student Engineers' Council and relevant student organizations (CSGSA, AWICS, TACS (TAMU ACM and IEEE student chapter), UPE and TAGD).
MANDATORY FOR NEW GRAD STUDENTS (but not other CSCE 681 students)
CSCE 681 Graduate Seminar (required for all new graduate students and all CSCE 681 students):
4:10-6:00 p.m., Wednesday August 29, 2012
Room 124, Bright Building
Abstract
MANDATORY FOR NEW GRAD STUDENTS and CSCE 681 STUDENTS
4:10 p.m., Monday, September 17, 2012
Room 124, Bright Building
Abstract
Future human space flight missions take astronauts deeper into space and require increased crew autonomy from Earth. Robots can perform hazardous or tedious tasks to reduce crew EVA and improve crew safety. Automated spacecraft systems can perform routine tasks to reduce crew workload and increase crew autonomy. Increased use of automation represents a fundamental change for human space flight. It is important to understand the performance implications of this change. Debra Schreckenghost/TRACLabs is leading an investigation of technology for automatically measuring human-automation performance while operations are in progress. Technology has been developed to monitor actions as they are performed, compute metrics in real-time, and makes these metrics available remotely via the web. Ms. Schreckenghost will describe the software for computing performance measures in real time and present results from evaluating this technology during NASA analog operations, including remote control of rovers during robotic field tests and simulated crew operations with procedure automation.
Related Papers
1. Debra Schreckenghost, Tod Milam, an Terrence Fong, "Measuring Performance in Real Time during Remote Human-Robot Operations with Adjustable Autonomy," IEEE Intelligent Systems, Vol. 25, No. 5, September, 2010, pp. 36-44.
2. Debra Schreckenghost, R. Peter Bonasso, David Kortenkamp, Scott Bell, Tod Milam and Carroll Thronesbery, "Adjustable Autonomy with NASA Procedures," in The 9th International Symposium on Artificial Intelligence, Robotics and Automation in Space (i-SAIRAS-08)
Biography
Debra Schreckenghost is a Senior Scientist with TRACLabs in Houston Texas, and the Associate Team Lead of the Human Factors and Performance Team at the National Space Biomedical Research Institute (NSBRI). She conducts research that combines her background in electrical engineering (MEE, Rice University) with her knowledge of NASA operations gained as a Shuttle flight controller. Her research interests include adjustable autonomy, software agents for the workplace, and human-automation interaction. She is PI of multiple projects investigating human interaction with robots and automation during complex dynamic operations for NASA, NSBRI, and DARPA.
Faculty Contact: Dr. Robin Murphy (murphy [at] cse.tamu.edu)
Dr. Fillia Makedon
Distinguished Professor and Department Chair
Director, the HERACLEIA Human Centered Computing Laboratory
Computer Science and Engineering Department
The University of Texas at Arlington
4:10 p.m., Monday, September 24, 2012
Room 124, Bright Building
Abstract
We present challenges in the design of patient-centered integrated computational frameworks to support rehabilitation and physical therapy. We focus on systems that can be used to monitor, diagnose and enhance physical activity and physical therapy compliance for persons with chronic neuromuscular disorders, where motor and cognitive functionalities are impacted. Neuromuscular disorders appear in many different conditions, including osteoarthritis, rheumatoid arthritis, cerebral palsy, autism, trauma, physical injury, aging, etc.
We describe a game-driven approach that combines the collection and fusion of game playing data (e.g., “exergaming”) with facial expressions, gestures, body motion, and contextual metadata from the person’s environment. The aim is tools that can learn and respond to a person’s needs and physical/emotional constraints. We describe tools developed for persons with dementia, cerebral palsy and Rheumatoid Arthritis (RA). As an example, we use RA which is a chronic systemic inflammatory disease attacking the joints. RA is a major cause of reduced quality of life. Preserving functional range of motion and enhancing cardiovascular health in persons with RA is a primary goal of physical therapy and is essential to ensuring maximal independence, ability to work and quality of life.
The system we are developing uses tools to monitor joint motions and motor performance, and merges with other physiological indicators. Computer vision, machine learning and other methods are used to quantify patient physical activity and provide appropriate incentives and rewards to the user for maintaining physical therapy and physical activity. New methods are being developed for accurate modeling and computing of human motion, gestures, and physical activity, along with approaches for automating the collection and processing of heterogeneous data streams, such as facial expressions to detect pain. This is joint work with an interdisciplinary team of experts at UTA and Northeastern University.
Biography
Fillia Makedon is Distinguished Professor and Department Head of Computer Science and Engineering at the U. of Texas at Arlington (UTA). She received her Ph.D. in Computer Science from Northwestern University in 1982. Between 1991-2006, she was professor of computer science at Dartmouth College where she founded and directed the Dartmouth Experimental Visualization Laboratory (DEVLAB). Prof. Makedon has received many NSF research awards in the areas of trust management, data mining, parallel computing, visualization, and knowledge management. She is author of over 300 peer-reviewed research publications. She directs the HERACLEIA Human Centered Laboratory (heracleia.uta.edu) that develops assistive technologies for human monitoring and smart health. She is member of several journal editorial boards and chair of the annual PETRA conference (www.petrae.org)
Faculty Contact: Dr. Nancy Amato (amato [at] cse.tamu.edu)
Dr. Nicolas E. Stier Moses
Associate Professor
Decision, Risk, and Operations Division
Columbia Business School
Columbia University
4:10 p.m., Wednesday, September 26, 2012
Room 124, Bright Building
Abstract
We consider a facility location game on a network in which consumers who are located on vertices wish to connect to the nearest facility. Knowing this, each competitor must choose a vertex to locate its facility, hoping to capture the largest-possible market share. The game resembles the classic Hotelling model of spatial competition. Focusing in a duopoly, we study the existence and the spatial location of pure-strategy Nash equilibria, for progressively more complicated classes of networks.
The case of trees is well-studied: equilibria are characterized by medians - nodes that minimize the distance to consumers, and hence always exist. For cycles, we find that an equilibrium exists when there is at least one vertex with a sufficiently-big demand, in which case it must also be a median. For more general graphs, we construct a tree of maximal bi-connected components and apply the results for trees and cycles to get sufficient conditions for equilibrium existence. This provides a complete and efficient characterization of equilibria for cactus graphs and other generalizations. As before, at equilibrium both competitors locate their facilities in medians.
The previous results generalize the classic study of Hotelling from a line to more complicated networks. Exploring this further, we establish a connection for additional classes of graphs such as quasi-median graphs, median graphs, Helly graphs, and strongly-chordal graphs, which include lattices and other naturally-occurring topologies in real-world networks. Furthermore, we establish a stronger result for strongly-chordal graph: as for trees, any median leads to an equilibrium.
Finally, we show that removing edges from a cactus increases consumer cost. This precludes situations like the Braess paradox, whereby removing an edge can improve the situation. This result implies that trees are worst-case networks because they are minimal instances with respect to inclusion. While equilibria can be arbitrarily inefficient with respect to centralized solutions, we quantify the resulting worst-case inefficiency with parametric upper bounds that depend on topological parameters of the network.
Biography
Professor Stier's research focuses on the impact that self-mindedness has on decentralized systems. In particular, his goal is to explore mechanisms that can help coordinate competitors, either by design or by offering the correct incentives. The main applications of his research are in supply chain management, and in distribution, transportation and telecommunication networks.
He received the 2006 Glover-Klingman Prize (best paper published in Networks) and the 2008 INFORMS Transportation Science and Logistics Section Best Paper Award. Prof. Stier teaches courses in MBA, Executive MBA and PhD programs about Managerial Decision Making and the interface between Operations and Game Theory. He received a Ph.D. degree from the Operations Research Center of the Massachusetts Institute of Technology.
Faculty Contact: Dr. Evdokia Nikolova (nikolova [at] cse.tamu.edu)
Dr. Tom Ball
Research Manager/Principal Researcher
Microsoft Research
Redmond, Washington
4:10 p.m., Wednesday, October 3, 2012
Room 124, Bright Building
Abstract
In the last decade, advances in satisfiability-modulo-theories (SMT) solvers have powered a new generation of software tools for verification and testing. These tools transform various program analysis problems into the problem of satisfiability of formulas in propositional or first-order logic, where they are discharged by SMT solvers, such as Z3 from Microsoft Research (MSR). In this talk, I'll review advances from MSR that expand the scope of SMT solvers on several fronts, including:
These advances are due to Nikolaj Bjorner, Leonardo de Moura, Ken McMillan, and Margus Veanes at MSR, and their colleagues and interns.
Biography
Dr. Tom Ball is a Principal Researcher and Research Manager at Microsoft Research (Redmond), where he works in the area of software engineering, having made contributions in program profiling, software model checking, and empirical software engineering. He is a 2011 ACM Fellow.
Faculty Contact: Dr. Gabriel Dos Reis (gdr [at] cse.tamu.edu)
Bernardus (Ben) Juurlink
Professor for Embedded Systems Architectures
Berlin University of Technology, Germany
4:10 p.m., Monday, October 8, 2012
Room 124, Bright Building
Abstract
Several recent works predict the future of multicore systems or identify scalability bottlenecks based on Amdahl's law. Amdahl's law implicitly assumes, however, that the problem size stays constant, but in most cases more cores are used to solve larger and more complex problems.
There is a related law known as Gustafson's law which assumes that runtime, not the problem size,is constant. In other words, it is assumed that the runtime on $p$ cores is the same as the runtime on $1$ core and that the parallel part of an application scales linearly with the number of cores.
We apply Gustafson's law to symmetric, asymmetric, and dynamic multicores and show that this leads to fundamentally different results than when Amdahl's law is applied. We also generalize Amdahl's and Gustafson's law and study how this quantitatively effects the dimensioning of future multicore systems.
Biography
Ben Juurlink is professor of Embedded Systems Architectures of the Electrical Engineering and Computer Science faculty of TU Berlin. He has an MSc degree from Utrecht University (NL) and a PhD degree from Leiden University (NL). In 1997-1998 he worked as a post-doctoral research fellow at the Heinz Nixdorf Institute in Paderborn (DE). From 1998 to 2009 he was a faculty member in the Computer Engineering laboratory of Delft University of Technology (NL).
His research interests include multi- and manycore processors, instruction-level parallel and media processors, low-power techniques, and hierarchical memory systems. He has (co-)authored more than 100 papers in international conferences and journals and received a best paper award at the IASTED PDCS conference in 2002. He has been the leader of several national projects, work package leader in several European projects, and is currently coordinator of the EU FP7 project LPGPU.
He is a senior member of the IEEE, a member of the ACM, and a member of the HiPEAC NoE. He served in many program committees, is area editor of the journal Microprocessors and Microsystems: Embedded Hardware Design (MICPRO), and is general co-chair of the HiPEAC 2013 conference.
Faculty Contact: Dr. Lawrence Rauchwerger (rwerger [at] cse.tamu.edu)
Dr. Lisa Anthony
Post-Doctoral Associate
Information Systems Department
University of Maryland Baltimore County
4:10 p.m., Monday, October 15, 2012
Room 124, Bright Building
Abstract
The field of Natural User Interaction (NUI) focuses on allowing users to interact with technology through the range of human abilities, such as touch, voice, vision and motion. Children are still developing their cognitive and physical capabilities, creating unique design challenges and opportunities for interacting in these modalities. This talk will describe Lisa Anthony's research expertise in understanding children's expectations and abilities with respect to NUIs and designing and developing new multimodal NUIs for children in a variety of contexts, including education, healthcare and serious games. Examples of projects she will present are recent work in understanding characteristics of children using touch and gesture interaction on mobile devices and past work in designing natural educational interactions for children.
Biography
Lisa Anthony is presently a post-doctoral research associate in the Information Systems Department at the University of Maryland Baltimore County (UMBC). She holds an MS in Computer Science (Drexel University, 2002), a second MS in Human Computer Interaction (Carnegie Mellon University, 2006), and a PhD in Human-Computer Interaction (Carnegie Mellon University, 2008). Her research interests include understanding how children can make use of advanced interaction techniques, with a special emphasis on pen and gesture interaction, for desktop and mobile applications in education, healthcare, and serious games. Her PhD dissertation topic investigated the use of handwriting input for middle school math tutoring software, and her simple and accurate multistroke gesture recognizer called $N is well-known in the field.
Faculty Contact: Dr. Tracy Hammond (hammond [at] cse.tamu.edu)
Dr. Gopal Gupta
Erik Jonsson Professor and Department Head
Department of Computer Science
University of Texas at Dallas
4:10 p.m., Wednesday, October 17, 2012
Room 124, Bright Building
Abstract
Coinduction is a powerful technique for reasoning about unfounded sets, unbounded structures, infinite automata, and interactive computations. Where induction corresponds to least fixed points semantics, coinduction corresponds to greatest fixed point semantics. In this talk I will give a tutorial introduction to coinduction and show how coinduction can be elegantly incorporated into logic programming to obtain the coinductive logic programming (co-LP) paradigm. I will also discuss how co-LP can be elegantly used for sophisticated applications that include (i) model checking and verification, including of hybrid/cyber-physical systems and systems specified using the pi-calculus (ii) planning and goal-directed execution of answer set programs that perform non-monotonic reasoning.
Biography
Gopal Gupta received his MS and Ph.D. in computer science from the University of North Carolina at Chapel Hill in 1987 and 1991 respectively, and his B. Tech. in Computer Science from IIT Kanpur in 1985. Currently he is the Erik Jonsson Chaired Professor and Head of Computer Science at the University of Texas at Dallas. His research interests are in logic programming, programming languages semantics/implementation, and assistive technology. He has published extensively in these areas. He serves as an area editor of the journal Theory and Practice of Logic Programming. He has served in the program committees of numerous conferences, and since January 2010, has served as the President of the Association for Logic Programming. His work on logic programming has been the basis of two startup companies.
Faculty Contact: Dr. Nancy Amato (amato [at] cse.tamu.edu)
Dr. Nicholas R. Gans
Assistant Professor
Department of Electrical Engineering
University of Texas at Dallas
4:10 p.m., Monday, October 22, 2012
Room 124, Bright Building
Abstract
Deploying autonomous robots and vehicles to search for certain objects has numerous applications, including surveillance, security, human/machine interaction, search and rescue, environmental monitoring and scientific exploration. In some cases, a specific object may be desired. In others, only basic descriptions may be given, such as color. The search task could be as vague as “find the most interesting object in the room.” I will discuss solving this task by casting it as an optimization problem. The first challenge is to describe the search task as an objective function that takes its maximum value when the desired object is in view. The second challenge is developing real-time optimization algorithms that move the robot to poses that systematically improve the objective function until it reaches its maximum. Ongoing challenges include overcoming local maxima, nonholonomic constraints and obstacles in the workspace or configuration space.
Biography
Dr. Gans is an Assistant Professor at The University of Texas at Dallas. His research interests include nonlinear and adaptive control, with focus on vision-based control and estimation, robotics and autonomous vehicles. Research projects have including vision-based mapping and localization, vehicle control, and target tracking. Current research includes autonomous sensors to maximize sensor information, human-machine interfaces, and use of novel vision sensors such as thermal and range cameras. Dr. Gans has published over forty peer-reviewed conference and journal papers, and he holds two patents in these areas.
Dr. Gans earned his M.S. in electrical and computer engineering and his Ph.D. in systems and entrepreneurial engineering from the University of Illinois at Urbana-Champaign in December 2005. Prior to joining UT Dallas, he worked as a postdoctoral researcher with the Mechanical and Aerospace Engineering Department at the University of Florida and as a postdoctoral associate with the National Research Council, where he conducted research on control of autonomous aircraft for the Air Force Research Laboratory Munitions Directorate and developed the Visualization Laboratory for simulation of vision-based control systems.
Faculty Contact: Dr. Nancy Amato (amato [at] cse.tamu.edu)
Dr. Daniel W. Goldberg
Assistant Professor
Geography
Texas A&M University
4:10 p.m., Wednesday, October 24, 2012
Room 124, Bright Building
Abstract
Spatial data are used in all aspects of research, practice, and policy-making ranging from location-based services for engaging potential customer-bases, to linking environmental exposures to human health outcomes, to planning for and responding to natural disasters. Unless obtained as a primary source through field work or direct observation, these data are often computationally-derived through processes such as natural language processing (NLP) of text documents or transcriptions, spatial integration of heterogeneous multi-modal data sources/sensors, or by the spatial interpolation of values over an area from a set of known observations. In all cases, quantitative representations of spatial accuracy and uncertainty are paramount to fitness-for-use estimations as well as for determining spatial confidence intervals for the results of subsequent operations or decisions made using these data.
This talk will explore the development of innovative computationally-based approaches for quantifying, representing, and reducing the uncertainty of spatial information derived through the process of geocoding, or translating location descriptions into digital geographic representations. This process involves a series of steps including NLP, probabilistic record-linkage, and spatial interpolation, among others, each of which produce and contribute non-stationary spatial error distributions independently and in concert to the resulting output. One goal of this research is to overcome the limitations of current metrics that describe the "quality" of geocoded data as qualitative classifications because these measures cannot be used by scientists or end-users in meaningful ways. Alternative measures are developed which rely on the quantification and propagation of uncertainty for each internal component of the geocoding process. Using these measures, algorithmic approaches for reducing spatial uncertainty will be presented which leverage spatial relationships, operators, and knowledge drawn from regional and local artifacts of the data available to a geocoding system.
Biography
Dr. Daniel W. Goldberg is an Assistant Professor of Geography at Texas A&M University. Dr. Goldberg holds BS, MS, and PhD degrees in Computer Science, the latter two from the University of Southern California and the former from Rutgers University in the Great State of New Jersey. He was most recently the Associate Director of the GIS Laboratory at USC. Dr. Goldberg’s research interests lie at the intersection of Computer Science and GIScience, a niche where he uses CS theory to advance GIS data structures, algorithms, and practices for applications within the Health and Social Sciences on topics ranging from environmental exposure assessment to the construction of techniques and systems for integrating and analyzing heterogeneous spatio-temporal data sources such as disease and information propagation. He has published widely on the development and application of geocoding techniques and is currently focusing on creating methods to compute and represent spatial uncertainty in geocoded data, and the means by which these uncertainties can be propagated to and within environmental health models.
Faculty Contact: Dr. Tracy Hammond (hammond [at] cse.tamu.edu)
Hye-Chung Kum
Research Assistant Professor
School of Social Work and
Department of Computer Science
University of North Carolina at Chapel Hill
4:10 p.m., Monday, October 29, 2012
Room 124, Bright Building
Abstract
Population Health informatics is an emerging research area at the intersection of public health, computer science, and statistics in which quantitative methods and computational tools are applied to big data about people to answer questions about our health. In health informatics, there is a constant need for record linkage, the process of identifying record pairs which belong to the same real-world entity, to integrate uncoordinated heterogeneous databases. Ambiguous links must be manually reviewed during approximate record linkage to enable accurate data integration. This requirement would seem to make it impossible for researchers to protect patients' privacy when reviewing health informatics data containing personally identifiable information (PII) under the trust mode. To address this problem, we offer a novel decoupled data system that blocks sensitive attribute disclosure via encryption. In addition, we evaluate three methods - (1) Chaffing, (2) Display control for clerical review, and (3) Manipulation of universe around the data - which, when used in combination, are effective in minimizing identity disclosure.
Biography
Dr. Hye-Chung Kum is a research associate professor in the School of Social Work with a joint appointment in the Department of Computer Science at the University of North Carolina at Chapel Hill (UNC-CH). She received her Ph.D. (2004) in Computer Science and M.S.W. (1998) from UNC-CH. Dr. Kum is a lead researcher of the Population Informatics Research Group with extensive research on sequential pattern mining, Datamining and Knowledge Discovery (KDD), digital government, population health informatics, privacy, and the use of KDD technology on administrative data for program evaluation and policy analysis through a long standing collaboration with NC-DHHS (Department of Health and Human Services).
Faculty Contact: Dr. Yoonsuck Choe (choe [at] cse.tamu.edu)
4:10 p.m., Wednesday, October 31, 2012
Room 124, Bright Building
Abstract
Hosted by Aggie Women in Computer Science http://awics.cse.tamu.edu/trickorresearch.php, Trick-or-Research will be a presentation of the research currently going on in the Computer Science and Engineering Department and an opportunity for undergraduates and new graduate students to tour the labs in the department. Join us if you are interested in research or just want to learn more about what is going on in our department. This is a two part event:
- Faculty Research Presentations: Each professor will be given 90 seconds to explain his or her research focus during the regular 681 seminar on Wednesday from 4:10-5:25 p.m. in 124 HRBB.
- Lab Tours: The research labs in the department will be open for visiting from 2:30-4:30 p.m. on Thursday, November 1. This is the Trick-or-Research part! Each lab you visit will have goodies to give out and give you a tour of what is going on in that lab. Lab passports will be provided and students who get at least 5 stamps in their passports will be eligible for a raffle for great prizes that will be held after the event.
Feel free to come and go if your schedule does not permit you to attend the entire event.
Faculty Contact: Dr. Nancy Amato (amato [at] cse.tamu.edu)
Dr. Rada Mihalcea
Associate Professor
Department of Computer Science & Engineering
University of North Texas
4:10 p.m., Monday, November 5, 2012
Room 124, Bright Building
Abstract
With more than 10,000 new videos posted online every day on social websites such as YouTube and Facebook, the internet is becoming an almost infinite source of information. One crucial challenge for the coming decade is to be able to harvest relevant information from this constant flow of multimodal data. In this talk, I will introduce the task of multimodal sentiment analysis, and present a method that integrates linguistic, audio, and visual features for the purpose of identifying sentiment in online videos. I will first describe a novel dataset consisting of videos collected from the social media website YouTube and annotated for sentiment polarity. I will then show, through comparative experiments, that the joint use of visual, audio, and textual features greatly improves over the use of only one modality at a time. Finally, by running evaluations on datasets in English and Spanish, I will show that the method is portable and works equally well when applied to different languages.
This is joint work with Veronica Perez-Rosas and Louis-Philippe Morency.
Biography
Rada Mihalcea is an Associate Professor in the Department of Computer Science and Engineering at the University of North Texas. Her research interests are in computational linguistics, with a focus on lexical semantics, graph-based algorithms for natural language processing, and multilingual natural language processing. She serves or has served on the editorial boards of the Journals of Computational Linguistics, Language Resources and Evaluations, Natural Language Engineering, Research in Language in Computation, IEEE Transactions on Affective Computing, and Transactions of the Association for Computational Linguistics. She was a program co-chair for the Conference of the Association for Computational Linguistics (2011), and the Conference on Empirical Methods in Natural Language Processing (2009). She is the recipient of a National Science Foundation CAREER award (2008) and a Presidential Early Career Award for Scientists and Engineers (2009).
Faculty Contact: Dr. Nancy Amato (amato [at] cse.tamu.edu)
Dr. Valentina Salapura
Master Inventor; Cloud Computing
IBM Thomas J. Watson Research Center
Yorktown Heights, NY
4:10 p.m., Wednesday, November 7, 2012
Room 124, Bright Building
Abstract
Driven by the need to reduce the total cost of ownership of increasingly more demanding workloads, cloud computing is being rapidly adopted across the marketplace. Within companies, private clouds are offering a more efficient way to manage and use private data centers.In the broader marketplace, public clouds offer the promise of buying computing capabilities based on a utility model, buying compute resources on demand and billed by usage rather than requiring capital expenditures that must be planned years in advance.
To deliver on these promises, cloud computing relies heavily on virtualization techniques and standardized environments to simplify provision and administration. As systems users are racing to move to the cloud, customer expectations on cloud availability and the cost of computing in a cloud are radically changing system requirements and data center operation.
For system design, virtualization efficiency and modularity in provisioning are becoming of paramount importance. Metrics such as virtual machines per core, per socket and per Watt are the new metrics that must be optimized for. Where dividing scale-up systems into multiple virtual machines was a previous sweet spot for optimizing virtual machine management, cloud computing must address the creation, operation and migration of virtual machines across system boundaries.
The standardization, virtualization, modularity and cross-system management capabilities also offer unique opportunities to provide system resilience: as virtual machines are becoming first class citizens, resilience techniques can build on a well-defined framework for providing recovery measures for replicating unresponsive services, and recovering failed services to respond to disaster scenarios.
Biography
Dr. Valentina Salapura is an IBM Master Inventor and System Architect at the IBM T.J. Watson Research Center where she is helping IBM realize the value of cloud computing in the Services Innovation Lab. Previously, Dr. Salapura helped define IBM’s future research agenda and strategy working with the worldwide IBM research organizations. Dr. Salapura was an architect and technical leader for the Power8 processor definition, and for the Blue Gene program since its inception. She has contributed to the architecture and implementation of several generations of Blue Gene Systems focusing on multiprocessor interconnect and synchronization and multithreaded, multicore architecture design and evaluation. Dr. Salapura is a recipient of the 2006 ACM Gordon Bell Prize for Special Achievements for the Blue Gene/L supercomputer and Quantum Chromodynamics. She is the author of over 60 papers and over 80 patents on processor architecture and high-performance computing. She received her PhD degree from Technische Universität Wien in Vienna, Austria. Dr. Salapura is an IEEE Fellow, an ACM Distinguished Speaker and a member of the IBM Academy of Technology.
Faculty Contacts: Dr. Nancy Amato (amato [at] cse.tamu.edu) and Dr. Lawrence Rauchwerger (rwerger [at] cse.tamu.edu)
Dr. Carlos Monroy
Game Research and Development
Center for Technology in Teaching & Learning
Rice University
4:10 p.m., Monday, November 12, 2012
Room 124, Bright Building
Abstract
STEMscopes is an on-line K-12 science curriculum developed at Rice University and presently serving almost one million students and fifty thousand teachers. STEMscopes users continuously generate and update a massive and rich dataset. This repository depicts usage of an inquiry-based science program, offering exciting research opportunities, with the ultimate goal of improving teaching and learning. The growing amount of data, almost 500 million items of activity in six months, poses challenges including data storage, complex aggregation and analysis with broader implications for pedagogy, big data, and learning. In this talk I will describe the curriculum and the underlying pedagogical framework, followed by a discussion on the problems we tackle, along with examples of visualizations based on existing data. Given the large amount of data amassed on a daily basis, I will discuss the use of methods from machine learning, data mining, linguistics computing and visualization science for making sense of and contextualize data.
Biography
Carlos Monroy received his Ph.D. in Computer Science from Texas A&M in 2010. His research interests include big data, learning analytics, digital libraries, educational games, and science dissemination. In his dissertation, Carlos improved information contextualization in the reconstruction of ancient sunken ships by exploiting relations in a multilingual glossary and an ontology. In the spring of 2009, he joined the Center for Technology in Teaching and Learning at Rice University in Houston Texas, where he works in research and development of games for education. In the fall of 2010 he successfully embedded a forensic science game based on the TV show CSI, into a virtual world geared toward teenagers (whyville.net). Presently, he conducts research on learning analytics and information visualization on a K-12 STEM online curriculum that serves nearly one million students and fifty thousand teachers in Texas.
Faculty Contact: Dr. Richard Furuta (furuta [at] cse.tamu.edu)
Dr. Barrett R. Bryant
Professor and Chair
Department of Computer Science & Engineering
University of North Texas
4:10 p.m., Wednesday, November 28, 2012
Room 124, Bright Building
Abstract
There are many problems whose solutions take the form of patterns thatmay be expressed using grammars (e.g., speech recognition, textprocessing, genetic sequencing, programming language development, etc.).Construction of these grammars is usually carried out by computer scientists working with domain experts. Grammar inference (GI) is the process of learning a grammar from examples, either positive (i.e., the pattern should be recognized by the grammar) and/or negative (i.e., the pattern should not be recognized by the grammar). This talk will present the application of grammar inference to software engineering, including recovery of domain-specific language (DSL) specifications from example DSL programs and recovery of a meta model from instance models which have evolved independently of the original meta model.
Biography
Barrett R. Bryant is Professor and Chair of Computer Science and Engineering at the University of North Texas. He received his B. S. in computer science from the University of Arkansas at Little Rock in 1979 and his Ph. D. in computer science from Northwestern University in 1983, after which he joined the University of Alabama at Birmingham, where he remained until 2011. He has held visiting appointments at a number of institutions, including Ibaraki University, Hitachi, Japan, the Naval Postgraduate School, Monterey, California, USA, and Tsinghua University, Beijing, China. His research interests include theory and implementation of programming languages, formal specification of software systems, and component-based software engineering, and he has authored or co-authored over 140 published papers in these areas. He serves on the Steering Committee of SAC (ACM Symposium on Applied Computing), and was formerly Chair of the ACM Special Interest Group on Applied Computing (2003-2009). He is a member of EAPLS, and a senior member of ACM and IEEE. Further details are available at http://www.cse.unt.edu/~bryant.
Faculty Contact: Dr. Nancy Amato (amato [at] cse.tamu.edu)
