| |
|
|
The Department of Computer Science
of Colorado State University, in cooperation with ISTeC
(Information Science and Technology Center), offers the CS
Colloquium series as a service to all who are interested
in computer science. When in-person meetings are possible, most seminars are scheduled for
Monday 11:00AM -- 11:50AM in CSB 130 or Morgan Library
Event Hall. For help finding the locations of our seminar
meetings, consult the on-line CSU campus map. |
Upcoming Events
CS Colloquium Schedule, Fall 2021
|
|
Computer Science Department Colloquium CS Faculty Rapid-Fire Presentations of Current Research: Q&A, first session Speaker: Computer Science Faculty, Colorado State University When: 11:00AM ~ 11:50AM, Monday August 30, 2021  Abstract: CS Faculty answers questions about their research, and opportunities in their research group for students at CSU. |
|
|
|
Computer Science Department Colloquium CS Faculty Rapid-Fire Presentations of Current Research: Q&A, second session Speaker: Computer Science Faculty, Colorado State University When: 11:00AM ~ 11:50AM, Monday September 13, 2021 Abstract: CS Faculty answers questions about their research, and opportunities in their research group for students at CSU. |
|
|
|
Computer Science Department Colloquium Learning Analytics Visualizations as Pedagogical Tools to Support Students’ Self-Regulated Learning Speaker: Marcia Moraes, Scholar, Department of Computer Science, and James E. Folkestad, Professor & University Distinguished Teaching Scholar, School of Education, Colorado State University When: 11:00AM ~ 11:50AM, Monday September 20, 2021 Abstract: Learning analytics (LA) emerged as an independent area from the field of academic analytics in 2010, and papers related to LA draw on a diverse range of literature from fields such as education, technology and social sciences. This diversity reflects in the works that have been done, such as predicting students’ academic success, identifying students that are at risk of failing their courses, detecting learning strategies used by students, and providing personalized feedback to students. Learning analytics visualizations have been used as a way to provide personalized formative feedback to support students’ self-regulated learning (SRL). In this talk I will present a short overview on student facing learning analytics visualizations and the importance of having a pedagogical theory to support the design and development of those visualization. Then, I will present a project that I am working in partnership with the Center for the Analytics of Learning and Teaching (C-ALT) here at CSU named U-Behavior. U-Behavior is a Teaching and Learning Tool and Method that provides students with a specific kind of learning analytics visualization called Retrieval Practice Activity (RPA) Graph, which represents an individual student study behavior. With those visualizations, students are aware of their study behaviors, are prompt to reflect on that, and to plan for future practice that will improve their long-term learning. Bio: Marcia Moraes is a Computer Science Scholar at Colorado State University. She received her Ph.D. in Computer Science in 2004 from Federal University of Rio Grande do Sul (UFRGS), Brazil, was an Assistant Professor at Pontifical Catholic University of Rio Grande do Sul (PUCRS) (2000-2016) in the Computer Science Department, Brazil, a researcher (2017-2019) and a post-doctoral fellow (2019-2021) in the Center for the Analytics of Learning and Teaching (C-ALT) at Colorado State University. She is also a Ph.D. student in the Education Science program, School of Education at Colorado State University. During her 21 years of experience in Higher Education, she taught on-campus and online courses at undergraduate and specialization levels, coordinate online courses, worked at the Academic Vice President’s Office for Undergraduate Studies, and did research on educational technology applied to learning and teaching. She had projects funded by Brazilian agencies such as FAPERGS, CNPq, and CAPES. She is a member of the Computer Science Education Group in the Computer Science Department, Colorado State University. Her research interests are in the area of technology enhancing learning and teaching, more specifically learning analytics and computer science education. James Folkestad is a professor in the School of Education at Colorado State University (CSU) and a faculty member within the Organizational Learning, Performance, and Change (OLPC) Program. He earned his Ph.D. in Educational Human Resource Development (EHRD) from Texas A&M in 1996. He is dedicated to the scientific investigation of how technology can be used to enhance learning, training, and innovative practice. In addition, he is the Director of the Center for the Analytics of Learning and Teaching (C-ALT) a research center at CSU. C-ALT is dedicated to advancing the use of analytics to drive teaching and learning innovation. |
|
|
|
Computer Science Department Colloquium An Intimate Look at the Fort Collins Power Grid Speaker: Jerry Duggan, Research Associate, Energy Institute at Colorado State University When: 11:00AM ~ 11:50AM, Monday September 27, 2021 Abstract: As part of the NSF Sustainability Research Network, CSU has been putting together a multi-dimensional view of the Fort Collins Power Grid. This view includes detailed electrical system topologies, Advanced Metering Infrastructure data, and substation-level billing data. Linking electrical endpoints to Larimer County property records provides the basis for categorizing loads by property type. Other data dimensions include location and sizes of solar installations, weather data, and socio-economic data by way of the U.S. census. Research with this data includes technical analysis of Fort Collins electric grid, technical and social comparisons of the Fort Collins system with other municipalities in the network, and impacts on grid-connect policies. Recent analyses provide insights into the impact of the COVID quarantine shutdown on electrical load in the city. This seminar will describe the data set, challenges with working with such large, multi-dimensional data, and details on specific projects underway. Bio: Jerry Duggan has been working the the CSU Energy Institute for the past four years, focusing on methane emissions modelling, energy network control systems, microgrids, and low-cost sensor networks. Jerry has 24 years of industrial experience, and holds an MS in Computer Science. |
|
|
|
Computer Science Department Colloquium Modeling and responding to internal cognitive states in real-time: What I’ve learned so far Speaker: Caitlin Mills, Assistant Professor, University of New Hampshire When: 11:00AM ~ 11:50AM, Monday October 4, 2021 Abstract: Although human-computer interaction systems are ubiquitous at this point, there is ample room for improvement -- particularly in domains that require an accurate representation of the user themselves: what they know, what they want, what they are feeling. One area that has received growing attention in recent years is how to model and effectively respond to internal states that are quite covert, such as attention, cognitive load, and emotions. This information allows a system to make a better estimate about how and when to respond. For example, you may zone out while reading this abstract without even knowing it yourself. In this talk, I will present an overview of my recent work that has focused on modeling user’s attention and cognitive load from multi-modal data channels (eye-gaze, pupillometry, EEG, log data, keystrokes, etc.) and across different tasks (reading, film, etc.). I will also discuss recent intervention systems (e.g., the Eye Mind Reader) that respond to such internal states in real-time, along with the considerations for designing systems that focus on users’ prolonged use. I will conclude with some future directions that include modeling new constructs (i.e., insights, rumination) and in different domains (driving, VR, conversations). Bio: Caitlin Mills is an assistant professor of psychology at the University of New Hampshire. She earned a masters (2014) and Ph.D. from the University of Notre Dame (2016) in the Emotive Computing Lab with Sidney D’Mello. Dr. Mills then completed a two-year postdoc with a focus on Cognitive Neuroscience at the University of British Columbia. Her research focuses modeling and responding to constructs related to mind wandering and affect. Dr. Mills’s interdisciplinary research program incorporates theoretical and methodological approaches from cognitive psychology, computer science, cognitive neuroscience and education. She is currently pursuing three main lines of research: 1) characterizing when mind wandering occurs and how it influences learning; 2) building machine learning detectors that can detect respond when someone goes off-task in real-time and designing effective interventions; and 3) conducting studies to uncover the dynamic brain network interactions that give rise to spontaneous thought. Dr. Mills has published over 65 papers including high impact journals (Human Computer Interaction, Proceedings of the National Academy of Sciences, Scientific Reports) and highly competitive conference proceedings (ETRA, UMAP, HFES) and holds multiple grants to support her work. |
|
|
|
Computer Science Department Colloquium What’d you Say? How the Properties of Language are Critical for the Development of Dynamic, Adaptive System Speaker: Laura Allen, Assistant Professor, University of New Hampshire When: 11:00AM ~ 11:50AM, Monday October 11, 2021 Abstract: The development of skills and strategies across domains requires individuals to have ample opportunities for deliberate practice with formative feedback. Automated, adaptive training systems have been developed as a means to offer such training without the reliance on expert professionals or tutors. These systems have often targeted well-defined domains -- in other words, they have focused on skills that are relatively easy to assess (e.g., mathematics, perception). More recently, however, research has turned to focus on the development of systems that can train more complex, ill-defined domains, such as writing and philosophy. Such endeavors are fruitful but require more complex methods for automated assessment. Natural Language Processing (NLP) can provide some solutions to this issue. In this talk, I will describe multiple IES-funded projects, which focus on the development of adaptive training systems that center around skills and strategies for ill-defined domains. I will describe how NLP can be used to develop explicit and stealth forms of assessment in such systems, as well as how it can be used to drive adaptivity and feedback to users. I will then turn to a discussion of current and future directions that focus on training in rapidly changing environments, such as misinformation processing and social media communication. Bio: Dr. Laura K. Allen is an Assistant Professor of Psychology at University of New Hampshire. She earned a B.A. in English Literature and Foreign Languages from Mississippi State University (2010), followed by a M.A. (2015) and Ph.D. in Psychology (Cognitive Science) from Arizona State University (2017). The principal aim of Dr. Allen’s research has been to theoretically and empirically investigate the higher-level cognitive skills that are required for successful text comprehension and production, as well as the ways in which performance in these domains can be enhanced through strategy instruction and training. This line of research has been accompanied by a second line of work that explores how technologies can be developed and leveraged to facilitate learning and training. The overall goal of this research is to develop technologies and computational methodologies that will have a broad impact on current practices in research and instruction across multiple dimensions. |
|
|
|
Computer Science Department Colloquium Liveness Analysis, Modeling, and Simulation of Blockchain Consensus Algorithms’ Ability to Tolerate Malicious Miners Speaker: Amani Altarawneh, Computer Scholar, Colorado State University When: 11:00AM ~ 11:50AM, Monday October 18, 2021 Abstract: The blockchain technology revolution and associated use of blockchains in various applications have resulted in many organizations and individuals developing and customizing their own fit-for-purpose consensus algorithms. Because security and performance are principally achieved through the chosen consensus algorithm, the reliability and security of these algorithms must be both assured and tested. This talk provides a methodology to assess such algorithms for their security level and performance is required; liveness for permissioned blockchain systems is evaluated. It focuses on permissioned blockchains because they retain the structure and benefits afforded by the blockchain concept while end users maintain control over their processes, procedures, and data. Thus, end users benefit from blockchain technology without compromising data security. The developed methodology is used to provide a liveness analysis of byzantine consensus algorithms for permissioned blockchains. Bio: Dr. Amani Altarawneh received her BA from Mu'tah University in Computer Science, Alkarak, Jordan, and MS in CS from Bridgewater State University in MA, USA. She received her Ph.D. in Computer Science ⁄ Cyber Security from the University of Tennessee at Chattanooga (UTC), Tennessee, USA. Her primary interests are cybersecurity, Blockchain, IoT & smart cities, and parallel & distributed systems. She worked on and participated in several projects that produced innovative techniques to evaluate security for blockchain systems, and IoT & smart cities applications. |
|
|
|
Computer Science Department Colloquium Deep Learning at the Speed of Light Speaker: Sudeep Pasricha, Professor and Chair of Computer Engineering, Dept. of Electrical and Computer Engineering, Colorado State University When: 11:00AM ~ 11:50AM, Monday October 25, 2021 Abstract: The massive data deluge from mobile, IoT, and edge devices, together with powerful innovations in data science and hardware processing, have established deep learning as the cornerstone of modern medical, automotive, industrial automation, and consumer electronics domains. Domain-specific deep learning accelerators such as Google’s TPU, Apple’s Bionic, and Intel’s Nirvana, now dominate CPUs and GPUs for energy-efficient deep learning processing. However, the evolution of these electronic accelerators is facing fundamental limits due to the slowdown of Moore’s law and the reliance on metal wires, which already severely bottleneck computational performance today. Silicon photonics represents a promising post-Moore technological alternative to overcome these limitations. Not only can photonic interconnects fabricated in CMOS-compatible processes provide near speed of light transfers at the chip-scale, but photonic devices can now also perform computations entirely in the optical domain. In this talk, I will present my vision of how silicon photonics can drive an entirely new class of sustainable deep learning hardware accelerators that can provide orders of magnitude energy improvements over today’s accelerators. I will discuss the evolution of silicon photonics over the past two decades, from integrated optics to photonic devices that can now be fabricated with low-cost CMOS-compatible manufacturing techniques. I will then cover new directions in power minimization, variation tolerance, fault resilience, and security for communication and computation with silicon photonics. I will share experiences from my journey over the past decade and a half towards the goal of realizing viable silicon photonic networks and computing substrates. I will end the talk with a discussion of the challenges and opportunities to achieve unparalleled energy-efficiency and performance gains in future manycore computing platforms with silicon photonics. Bio: Sudeep Pasricha received the B.E. degree in Electronics and Communication Engineering from Delhi Institute of Technology, India, in 2000, after which he spent several years working for STMicroelectronics, India ⁄ France, and Conexant, USA. He received his Ph.D. degree in Computer Science from the University of California, Irvine in 2008. He joined Colorado State University in 2008 where he is currently a Walter Scott Jr. College of Engineering Professor in the Department of Electrical and Computer Engineering. He is a former University Distinguished Monfort Professor and Rockwell-Anderson Professor. He is currently also Chair of Computer Engineering and Director of the Embedded, High Performance, and Intelligent Computing (EPIC) Laboratory at Colorado State University. His research broadly focuses on software algorithms, hardware architectures, and hardware- software co-design for energy-efficient, fault-tolerant, real-time, and secure computing, for embedded, IoT, and cyber-physical systems. Prof. Pasricha has published more than 200 papers in peer-reviewed journals and conferences that have received seven best paper awards and six best paper nominations. He has filed for multiple patents and co-authored several books and book chapters. He has also given several keynotes, invited talks, and tutorials. His contributions have been recognized with several awards, including the George T. Abell Outstanding Research Faculty Award, IEEE-CS ⁄ TCVLSI Mid-Career Research Achievement Award, IEEE ⁄ TCSC Award for Excellence for a Mid-Career Researcher, AFOSR Young Investigator Award, ACM Technical Leadership Award, and ACM SIGDA Distinguished Service Award. He is currently the Vice Chair of ACM SIGDA and the Steering Committee Chair for the IEEE Transactions on Sustainable Computing. He is also a Senior Associate Editor for the ACM Journal of Emerging Technologies in Computing, and an Associate Editor with several ACM and IEEE journals. He has served as General Chair and Technical Program Chair of 12 conferences, Steering and Organizing Committee Member of 40+ conferences, and Technical Program Committee Member of 100+ conferences. He is an IEEE Senior Member and an ACM Distinguished Member. |
|
|
|
Computer Science Department Colloquium IoT and Mobile Crowdsourcing: Enablers of Smart Cities Speaker: Qi Han, Professor, Department of Computer Science, Colorado School of Mines When: 11:00AM ~ 11:50AM, Monday November 1, 2021 Abstract: In recent years, Internet of Things (IoT) has gained increasing attention and wider adoption in various applications. At the same time, human being nowadays often carry different mobile devices such as wearables and smartphones with built-in cameras, so the crowd may be recruited to accomplish some tasks. These technology advances have made cities smarter than ever before. I will use specific projects to illustrate how we address challenging research issues in order to unleash the power of the two key enablers (i.e., IoT and mobile crowdsourcing) of smart cities. I will first discuss multi-layer adaptive techniques we have developed for quality-aware voice stream multicast over multi-hop low power wireless networks for effective search and rescue. I will then present several building blocks for mobile visual crowdsensing and sharing systems. I will end the talk with a discussion of the challenges and opportunities to achieve true smart cities based on insights gained from my journal over the past decade. Bio: Qi Han (http: ⁄ ⁄ www.mines.edu ⁄ ~qhan) is Professor in the Department of Computer Science at the Colorado School of Mines. She founded and currently directs the Pervasive Computing Systems (PeCS) research group (http: ⁄ ⁄ pecs.mines.edu). Her broad research interests lie in the areas of pervasive computing and mobile systems, with current focus on applying mobile sensing, crowdsourcing, Internet of Things, swarm robotics, and real-time analytics to understand human activities and improve safety and efficiency of human life. She has also been active in interdisciplinary research where she applies her expertise to a variety of applications in the domains of smart cities and space exploration. Her research has been mainly funded by the National Science Foundation (NSF), National Aeronautics and Space Administration (NASA), and Army Research Labs (ARL). Dr. Han holds a Ph.D. degree from the Donald Bren School of Information and Computer Sciences at the University of California, Irvine. She has served on a number of technical program committees for international conferences and held several workshop or conference program chair positions. She is an ACM Distinguished Speaker, an ACM senior member, and an IEEE senior member. |
|
|
|
Computer Science Department Colloquium Deception and misinformation across natural language genres Speaker: Ritwik Banerjee, Research Assistant Professor, Computer Science, Stony Brook University When: 11:00AM ~ 11:50AM, Monday November 29, 2021 Abstract: Identifying misinformation requires contrasting a claim with established relevant facts. Whether this identification is done through computational means or human acumen, it requires encyclopaedic knowledge about a larger world beyond a single piece of text. Further, it requires a background belief system and trust. In this talk, I will discuss how this trust can be manipulated (intentionally or otherwise) to deceive the recipient of a piece of communication, and how this contributes to the spread of misinformation. Specifically, I will delve into information in the domain of healthcare and medicine that travels across genres, where the style as well as the audience of communication undergo drastic changes, and discuss their implications. I will conclude the talk with a breakdown of the current challenges in identifying misinformation and deception in natural language, signalling avenues for future research. Bio: Dr. Ritwik Banerjee is a Research Assistant Professor of Computer Science at Stony Brook University. His research interests include knowledge discovery, natural language inference, and misinformation analysis. These have made him venture into applications in precision healthcare, such as prediction of adverse drug events in emergency rooms, and also into information extraction from financial corpora. His most recent work has been on empirical investigations into cross-genre misinformation, specifically in health news, delving into not just fake news in itself, but also the use of deceptive linguistic and extra-linguistic cues in their presentation. His research has been supported by the National Science Foundation (NSF) as well as by industry funding. Dr. Banerjee has served on a number of program committees for international conferences and workshops. He received his Ph.D. in Computer Science from Stony Brook University, preceded by a M.Sc. in Computer Science, and B.Sc. in Mathematics, both from the Chennai Mathematical Institute (India). |
|
|
|
Computer Science Department Colloquium Quo Vadis, Polyhedral Model Speaker: Sanjay Rajopadhye, Professor, Department of Computer Science, Colorado State University. When: 11:00AM ~ 11:50AM, Monday December 6, 2021 Abstract: Exponential growth of transistor density is ending, or at least diminishing, but Moore’s Law of societal expectation continues unabated. To deliver on it, we need seamless integration all across the stack of heterogeneous, accelerator-rich architectures, multi-core and embedded processors, through distributed systems and clouds. Emerging workloads are driven largely by the computational needs of AI ⁄ ML. The polyhedral model provides a powerful compiler infrastructure, applicable for most of them. It is now found in modern compilers and frameworks like llvm, and is the subject of very intense commercial investment with large research and development teams. But these efforts mostly target low-hanging fruit: polyhedral programs that are easy to parallelize. In this talk, I will argue that relatively small, single-investigator efforts should return to the fundamentals and tackle the harder and deeper problems. I will illustrate with two of them: (i) compiler techniques to automatically invent algorithms, and (ii) parallelization of inherently and apparently largely sequential programs. Bio: Sanjay Rajopadhye is currently professor in the Computer Science (CS) and in the Electrical and Computer Engineering (ECE) departments at Colorado State University. He is one of the inventors of the polyhedral model. His Ph.D. dissertation made contributions to (i) scheduling and its limitations, (ii) closure properties, and (iii) data reuse. All topics are increasingly relevant for post-Moore computation. |
