RESEARCH KEYNOTE SERIES
(Professor, University of Windsor)
Bio: Prof. Rueda received his Bachelor’s degree in computer science from the National University of San Juan, Argentina, in 1993, and his Master’s and Ph.D. degrees in computer science from Carleton University, Canada, in 1998 and 2002, respectively. He is currently a Full Professor in the School of Computer Science at the University of Windsor. His current research interests are mainly focused on devising shallow and deep machine learning and representation learning algorithms at the fundamental level and applications in bioinformatics and cybersecurity to problems in protein-protein interaction, transcriptomics,
integrative genome-wide analysis, identification of cancer biomarkers, user authentication, spam review detection and social engineering. Luis Rueda holds four patents on cybersecurity and has more than 200
publications and presentations in prestigious journals and conferences in machine learning, computational biology and data security. He currently serves as Associate Editor of IEEE/ACM Transactions on Computational Biology and Bioinformatics. He is also a member of the Technical Committee on Pattern Recognition for Bioinformatics (IAPR TC-20) and the program committees of several conferences in the field. He is also a Senior Member of the IEEE, and a Member of the Association for Computing Machinery and the International Society for Computational Biology.
Title of talk: How Self-Organizing Maps Empower Convolutional Neural Networks
Abstract: One of the main challenges in representation learning is to deal with high-dimensional, unstructured data for classification, where the inherent complex relationship among points can be understood toward a representation into a significantly lower dimension, viz Euclidean spaces. The challenge becomes more prevalent when dealing with various types of heterogeneous data. This talk discusses a systematic, generalized method that uses self-organizing maps (SOMs) to represent higher dimensions onto a two- dimensional grid. Afterwards, a convolutional neural network (CNN) is applied to predict or classify samples such as documents or disease states of various types. Integration of multi-omics data for prediction of cancer subtypes and detection of biomarkers will be discussed. In addition, the talk will show the use of information about semantically- similar words around mapped via a SOM, followed by prediction of spam reviews via a CNN.
Bipin C. Desai
(Professor, Concordia University)
Bio: Dr. Desai, was one of the founding members of the first Computer Science department in Canada. Over the years he has worked on systems including control systems and aircraft simulation system including the visual simulators. He has been working in database systems and their applications; his recent interests are in Privacy and Security. He has been the General chair for the C3S2E and IDEAS series of international conferences.
Title of talk: Closed vs Open Systems
Abstract: The marketing of computing systems in the early days included the bundling of basic software support. This included the operating system, the compilers and libraries as well as training manuals. This model was subjected to a monopoly What has happened now is that closed systems are marketed today that have the software, including privacy compromising bugs and tracking sub-systems, built into them. All the software applications created by independent software houses are installed via the operating system of the closed system and the device maker imposes a percent charge. There is no move anywhere to un-bundle software, including the applications and the hardware. In this talk, we will examine the philosophy of moving away from such closed systems to improve innovation and hence create better systems.
Peter van Beek
(Professor, University of Waterloo)
Bio: Peter van Beek is a Professor in the Cheriton School of Computer Science at the University of Waterloo. He received his PhD in 1990 from the University of Waterloo and at that time joined the faculty at the University of Alberta. Ten years later he returned to Waterloo. His research interests span the field of artificial intelligence with a current focus on probabilistic graphical models, constraint programming, and applied machine learning. Peter has co-authored seven research papers which have won awards. In 2008, he was named a Fellow of the Association for Artificial Intelligence and in 2019, he was named a Fellow of the Canadian Artificial Intelligence Association.
Title of Talk: Data Analysis using Bayesian Networks
Abstract: Bayesian networks are widely used as a data analysis tool in diverse areas, including finance, medicine, and sports. In this talk, I will review Bayesian networks and discuss their use and advantages for data analysis. I will then focus on our recent work on learning the graphical structure of a Bayesian network from discrete data based on learning *all* near-optimal networks. Our approach allows us to identify features such as edges between variables with high confidence and has led to improved data analysis algorithms both in terms of accuracy and scale. The talk will be aimed at a general audience.
Ivan V. Bajić
(Professor, Simon Fraser University)
Bio: Ivan V. Bajić is a Professor of Engineering Science and co-director of the Multimedia Lab at Simon Fraser University, in Burnaby, BC, Canada. His research interests include signal processing and machine learning with applications to multimedia processing, compression, and collaborative intelligence. His group’s research has received awards at ICME 2012 and ICIP 2019, other recognitions (e.g., paper award finalist, top n%) at Asilomar, ICIP, ICME, and CVPR, and was featured in the IEEE Signal Processing Magazine (May 2020), the front page of the IEEE Transactions on Audio, Speech, and Language Processing (July/August 2016), among featured articles in the IEEE Transactions on Image Processing (volume 30, 2021), as well as popular media such as Vancouver Sun, Plank Magazine, and CBC Radio. He has received an NSERC DAS Award in 2021 for his work on collaborative intelligence. He was an Associate Editor of the IEEE Transactions on Multimedia and the IEEE Signal Processing Magazine, and is currently serving as a Senior Area Editor of the IEEE Signal Processing Letters.
Title of Talk: Error-Resilient Collaborative Intelligence
Abstract: Edge-cloud Collaborative Intelligence (CI) is a framework in which AI models are distributed between the edge devices and the cloud. Typically, the front-end of an AI model is deployed on an edge device, where it performs initial processing and feature computation. These intermediate features are sent to the cloud, where the back-end of the AI model completes the inference. CI has been shown to have the potential for energy and latency savings compared to the more typical cloud-based or fully edge-based AI model deployment. However, new challenges arise from the fact that there is now an error-prone channel in the information pathway of the distributed AI model. The focus of this talk is error resilience in CI systems. I will describe certain properties of intermediate features from deep neural networks that make error-resilient CI possible. I will also present an overview of recent work on tensor completion for recovery of missing data in packet-based transmission of deep feature tensors, along with promising avenues for future work.
(Professor, Simon Fraser University)
Bio: Sheelagh Carpendale is a Professor and Canada Research Chair in Information Visualization in the School of Computing Science at Simon Fraser University. She is a Fellow in the Royal Society of Canada has been inducted into the both IEEE Visualization Academy and the ACM CHI Academy. She has received many awards including the IEEE Visualization Career Award, an NSERC STEACIE (a top science award in Canada); a BAFTA (the British equivalent to an Oscar); an ASTech Award Innovations in Technology; two Best Supervision Awards; and the Canadian CHCCS Achievement Award. She leads the Innovations in Visualization (InnoVis) Research Group and has newly established the Interactive Experiences Lab (ixLab). Her research on information visualization, interaction design and qualitative empirical research draws on her dual background in Computer Science (BSc. and Ph.D.) and Visual Arts (Sheridan College, School of Design and Emily Carr, College of Art). By studying how people interact with information in both work and social settings, she works towards designing more inclusive, accessible and understandable interactive visual representations of data. She combines information visualization and human-computer interaction with innovative new interaction techniques to better support the everyday practices of people who are viewing, representing, and interacting with data.
Title of Talk: Data Visualization for Empowerment and Inclusion
Abstract: In this talk I touch on a few moments in my research history that have changed my way of thinking and gradually leading me to become passionate about – data visualization for empowerment and inclusion. I have chosen moments that are memorable to me because they have changed the way I think. I will step through my process of discovering the importance of thinking about how to make visualizations more empowering, more engaging, and more expressive, and how may lead to developing visualizations that work better for us as externalizations. I will relate these points to discussions and data feudalism and the importance of us staying connected to our data heritage.
(Professor, University of Alberta)
Bio: Ioanis Nikolaidis is a Full Professor and Associate Chair of the Department of Computing Science at the University of Alberta. He supervised several PhD, MSc students, as well as undergraduate and high school trainees and has published more than a hundred papers and chapters in refereed journals, conferences, and books. Dr. Nikolaidis is conducting research in the area of data networking protocols, protocol performance, and wireless sensor networks. He served as Area Editor for Computer Networks, Elsevier ('00-'10), and as a long-standing Editor ('99-'13) and Editor-in-Chief ('07-'09) for the IEEE Network magazine. He co-chaired CNSR 2011, ADHOC-NOW 2004 & 2010, and co-organized SSS 2015, and served as technical program committee member and reviewer for numerous conferences and journals, as well as a reviewer or panel member for many funding agencies: NSERC, NSF, OCE, FWF/START, NWO, etc. He co-edited the book "Building Sensor Networks: From Design to Applications" (CRC Press). He is a member of IEEE and a lifetime member of ACM.
Title of Talk: Challenges in Building Grass Roots Wireless Sensor Networks
Abstract: We will describe our experiences over the last two decades with building real systems of wireless sensors using off-the-shelf, and generally inexpensive, components. Even today, with a diverse set of sensor offerings and protocols, one does not always find a good fit for particular applications. Rethinking an application from the wireless sensor network point of view usually involves considerations related to: a) energy supply limitations, b) data transfer limitations, and c) protocol flexibility. At the same time, any solution has to fit within the meagre resources of a, generally, low end computational platform. For energy considerations, we will present protocol-level mechanisms and energy harvesting strategies. For data transfer limitations, we will outline the potential benefit of in-network processing, namely as various forms of data aggregation. Finally, for protocol flexibility, we will introduce an expressive rule-based cross-layer forwarding rule system, that captures a vast set of low-level protocol mechanisms while being resource conscious.
(Professor, University of Windsor)
Bio: Dr. Xiaobu Yuan is a full professor with the School of Computer Science at the University of Windsor, Canada. He started research on Virtual Reality (VR) twenty years ago, with his first paper on the subject published in 1997. For his contributions of applying VR in robotics and software engineering, he was invited to serve three times in the technical program of IEEE ICRA (IEEE International Conference on Robotics and Automation) and in program committees for more than 40 IEEE and other international conferences. His research mainly focuses on inventing new means for the creation of joint human-computer intelligence via VR, and his on-going research projects include the use of software avatar for interactive software customization and the construction of simulation environment for autonomous vehicles.
Title of Talk: Digitalization of Software Development via Virtual Reality
Abstract: In the new age of information technology, enterprises are following the trend of digital transformation to foster innovations by converting their operation models. While software product line engineering (SPLE) promises huge increase of productivity by means of mass production, the lack of support for interactive software customization has become one of the major obstacles that slows down advancements in software industry. This talk will first highlight the benefits of SPLE for developing a diversity of similar systems and then discuss the ability of VR for creating digital counterparts of real-word entities before exploring the potentials of digitalizing software development by allowing the virtual representation of software developers to assist software clients “ordering” software products online according to their need.
Eyal de Lara
(Professor, University of Toronto)
Bio: Dr. Eyal de Lara is a Professor in the Department of Computer Science at the University of Toronto. His focus is on experimental research on mobile and pervasive computing systems. Prof. de Lara served as editor in chief of ACM GetMobile the flagship publication of ACM SigMobile from 2014 to 2018. His research has been recognized with the 2019 EuroSys Test of Time Award, an IBM Faculty Award, an NSERC Discovery Accelerator Award, the 2012 CACS/AIC Outstanding Young Computer Science Researcher Prize, and two best paper awards.
Title of Talk: Storage Systems for the Edge
Abstract: The advent of the Internet of Things (IoT) has led to a large amount of generated data, new requirements for real-time processing, and the proliferation of distributed heterogeneous computing resources. In emerging scenarios, such as Smart Cities and Industry 4.0, files, videos, and pictures are produced by sensors on the edges of the Internet. Transporting this data over Internet links to cloud data centers for processing and storage places several bandwidth requirements on the network and is not compatible with applications that require low latency. Edge computing has emerged as an alternative approach that leverages small regional data centers, from a few computers to a few racks, at the Internet edges to carry out processing tasks and store data. This proximity to the location of data generation enables new systems that provide low latency and reduce network load. In this talk, I will describe recent research on distributed storage systems designed to simplify the deployment of novel applications on a hierarchy of data centers that span from the edge of the network for the wide-area cloud.
|Full Paper Submission:||20th September 2021|
|Acceptance Notification:||7th October 2021|
|Final Paper Submission:||15th October 2021|
|Early Bird Registration:||15th October 2021|
|Presentation Submission:||19th October 2021|
|Conference:||27 - 30 October 2021|
• Conference Proceedings will be submitted for publication at IEEE Xplore® digital library .
• Best Paper Award will be given for each track.
• Conference Record No 53756