Agenda

Title: Data Science, Artificial Intelligence, and Machine Learning

Abstract

The three concepts in the title are related and often confused with one another. We shall try to separate the concepts and then discuss the limitations of machine learning. We close with two examples of algorithms that are useful in data science but are not machine learning: locality-sensitive hashing and counting distinct elements.

Biography

Jeff Ullman is the Stanford W. Ascherman Professor of Engineering (Emeritus) in the Department of Computer Science at Stanford and CEO of Gradiance Corp. He received the B.S. degree from Columbia University in 1963 and the PhD from Princeton in 1966. Prior to his appointment at Stanford in 1979, he was a member of the technical staff of Bell Laboratories from 1966-1969, and on the faculty of Princeton University between 1969 and 1979. From 1990-1994, he was chair of the Stanford Computer Science Department. Ullman was elected to the National Academy of Engineering in 1989, the American Academy of Arts and Sciences in 2012, the National Academy of Sciences in 2020, and has held Guggenheim and Einstein Fellowships. He has received the Sigmod Contributions Award (1996), the ACM Karl V. Karlstrom Outstanding Educator Award (1998), the Knuth Prize (2000), the Sigmod E. F. Codd Innovations award (2006), the IEEE von Neumann medal (2010), the NEC C&C Foundation Prize (2017), and the ACM A.M. Turing Award (2020). He is the author of 16 books, including books on database systems, data mining, compilers, automata theory, and algorithms.

Title: Supercloud Computing

Abstract

A new era of hyperscale intelligent computing is emerging, driven by new cloud computing business models, new hardware for general-purpose and accelerated computing, and new hardware and software for infrastructure processing. There are, however, many major technical challenges involved in developing the future 'superclouds' required. We need new architectures that can continuously run many large-scale intelligent applications concurrently on a single shared infrastructure. The applications might involve HPC, AI, Big Data Analytics, Graph Computing or Knowledge Computing, but all will need to be delivered with the high performance, efficient resource utilization, and nonstop resilience required by commercial hyperscale services. Some applications will require not only that, but also to be able to perform HPC, AI, Analytics and Knowledge Computing together in a tightly integrated way within a single application.

What is now urgently needed are new theories and models that can provide a foundation to guide this new era. With such a theoretical foundation in place, we can begin to design the new architectures, algorithms, programming tools and software systems that we need. In this talk I will introduce a new model for supercloud computing that provides such a foundation for this next stage in the evolution of the computing industry.

Biography

Bill McColl is the Director of the Computing Systems Lab at Huawei's Zurich Research Center, where he leads fundamental research on future computing architecture, software and algorithms. He is also a Fellow of Wadham College, Oxford University. Previously he was Professor of Computer Science, Head of Research in Parallel Computing, and Chairman of the Faculty of Computer Science at Oxford. He established and led Oxford Parallel, a major center for research on industrial and business applications of HPC at the university. Much of his previous research was focused on the Bulk Synchronous Parallel (BSP) approach to parallel architecture, software and algorithms. BSP is now used throughout industry for massively parallel graph databases, graph analytics, machine learning and other areas of AI. His current research is focused on developing a new theoretical foundation and model for current and future hyperscale computing.

Title: Edge Intelligence

Abstract

As humans, Internet of Things, software services and AI continue to become the entangled fabric of distributed systems, systems engineers and researchers are facing novel challenges. In this talk, we analyze the role of AI in context of IoT, Edge, and Cloud in the co-evolution of distributed systems for the new decade. We identify challenges and discuss a roadmap that these new distributed systems have to address in order to bring intelligence to the edge. We take a closer look at how a cyber-physical fabric will be complemented by AI operationalization to enable seamless end-to-end distributed systems.

Biography

Schahram Dustdar is a Full Professor of Computer Science at the Vienna Technical University, heading the Research Division of Distributed Systems at the TU Wien, Austria. He holds several honorary positions:
University of California (USC) Los Angeles; Monash University in Melbourne, Shanghai University, Macquarie University in Sydney, University Pompeu Fabra, Barcelona, Spain. From Dec 2016 until Jan 2017 he was a Visiting Professor at the University of Sevilla, Spain and from January until June 2017 he was a Visiting Professor at UC Berkeley, USA.

From 1999 - 2007 he worked as the co-founder and chief scientist of Caramba Labs Software AG in Vienna (acquired by ProjectNetWorld AG), a venture capital co-funded software company focused on software for collaborative processes in teams. He is co-founder of edorer.com (an EdTech company based in the US) and co-founder and chief scientist of Sinoaus.net, a Nanjing, China based R&D organization focusing on IoT and Edge Intelligence.

He is founding co-Editor-in-Chief of ACM Transactions on Internet of Things (ACM TIoT) as well as Editor-in-Chief of Computing (Springer). He is an Associate Editor of IEEE Transactions on Services Computing, IEEE Transactions on Cloud Computing, ACM Computing Surveys, ACM Transactions on the Web, and ACM Transactions on Internet Technology, as well as on the editorial board of IEEE Internet Computing and IEEE Computer.

Dustdar is recipient of multiple awards: IEEE TCSVC Outstanding Leadership Award (2018), IEEE TCSC Award for Excellence in Scalable Computing (2019), ACM Distinguished Scientist (2009), ACM Distinguished Speaker (2021), IBM Faculty Award (2012). He is an elected member of the Academia Europaea: The Academy of Europe, where he currently is chairman of the Informatics Section, as well as an IEEE Fellow (2016) and an Asia-Pacific Artificial Intelligence Association (AAIA) Fellow (2021) and the AAIA president (2201).

Title: Autonomous Systems and AI

Abstract

We argue that, despite the spectacular achievements of neural networks, we are still far from AGI. A big step toward this goal would be to develop autonomous systems capable of replacing human agents working in complex organizations, as envisioned by the IoT. This requires in particular, the convergence between Computing and AI integrating data-based machine learning and model-based systems engineering.

We discuss the relevance of existing criteria for comparing human and machine intelligence and show some notable analogies and differences between scientific knowledge and that produced by neural networks.

Emphasizing that autonomy is an important step towards AGI, we present a characterization of autonomous systems, and show key differences with mental systems equipped with common sense knowledge and reasoning.

We conclude by advocating challenging work directions, including the development of a new foundation for systems engineering and scientific knowledge, and the joint exploration of physical and mental phenomena that embody human intelligence.

Biography

Joseph Sifakis was born in Heraklion, Crete in 1946 and lives in France. He studied Electrical Engineering at the National Technical University of Athens and Computer Science at the University of Grenoble under a French scholarship. He received his engineering doctorate in 1974 from the University of Grenoble, where he also received a state doctorate in 1979.

He is currently Research Director Emeritus for the Centre national de la recherche scientifique at VERIMAG laboratory near Grenoble, of which he is the founder.[circular reference] Sifakis has been a leading figure in the fields of Model Checking and Embedded Systems. He founded with Edmund M. Clarke and Amir Pnueli the CAV conference, organized for the first time in Grenoble in 1989. He has been the coordinator of the ARTIST European Network of Excellence for research on Embedded Systems(2004–2012).

Sifakis held the INRIA-Schneider endowed industrial chair (2008–2011) and has been a full professor and the Director of the Rigorous System Design Laboratory at the School of Computer and Communication Sciences of EPFL (2011–2016). He has been visiting professor at Tsinghua University (2011–2012) and is currently a visiting professor at SUSTech University.

Sifakis has been the President of the Greek National Council for Research and Technology (2014–2016).

Title: Reinventing Healthcare with Artificial Intelligence: drug development, diagnosis, treatment implementation and prognosis

Abstract

Artificial intelligence is a game changer in healthcare. The increasing availability of data / computing power along with progress in the field of machine learning / deep learning are transformative pillars towards optimization of treatment pathways. Artificial intelligence offers the ability to reason holistically on multi-modal data and establish causal relations between patient - prior to treatment - data and observed outcomes. This has been beneficial on addressing the entire healthcare landscape; starting from early stage clinical drug development, all the way to efficient treatment implementation. The aim of this presentation is to provide an overview of the potential impact of artificial intelligence in healthcare, the most commonly adopted scientific components and concrete examples of its effective implementation in the clinical workflow.

Biography

Nikos Paragios (50) is distinguished professor of Mathematics (on partial leave) at CentraleSupelec, the school of engineering of the University of Paris-Saclay and Chief Executive Officer of TheraPanacea.

Prior to that he was senior fellow (2015-2020) at the Institut Universitaire de France, director (2011-2017) of the Center for Visual Computing at CentraleSupelec, scientific leader (2007-2017) at Inria (Galen Research Team), professor (2004-2005, 2011-2013) at the Ecole de Ponts ParisTech, and affiliated with Siemens Corporate Research (Princeton, NJ, 1999-2004) as a project manager, senior research scientist and research scientist. N. Paragios was an adjunct professor at Rutgers (2002) and at New York University (2004) and a visiting professor at Yale (2007) and at University of Houston (2009).

Professor Paragios is a fellow of IEEE (class of 2011), has co-edited four books, published more than 350 papers in the most prestigious journals and conferences (DBLP server), and holds 30 international patents. His work has approx 27,500 citations according to Google Scholar and his H-number (06/2022) is 72. Professor Paragios is the Editor in Chief of the Computer Vision and Image Understanding Journal (CVIU - IF 3.88) since 2013 and serves as member of the editorial board for the Medical Image Analysis Journal (MedIA) while has served in editorial board of numerous and prestigious international journals (IEEE T-PAMI, IJCV, IVC, MVA, JMIV, SIIMS).

Pr. Paragios received an IBM faculty award in 2014, a European Research Council excellence award in 2011 while in 2008 was the laureate of the Greece's (nationality or descent) highest honor (world-wide) for young academics and scientists, the Bodossaki Prize. In 2006, he was named one of the top 35 innovators world-wide innovators under the age of 35 from the MIT's Technology Review. Professor Paragios is member of the scientific council of Safran conglomerate.

Title: From Imaging and 3D Vision to AI based Brain Connectomics

Abstract

This talk is an invitation to a journey starting from Computational Imaging and 3D Computer Vision to AI based Brain Connectomics. I emphasize the role and power of Modelling, Geometry and AI in solving some of the most challenging problems I contributed over the last years, for processing multi-valued images, building 3D models and mapping the human brain connectivity. Variational methods and geometric flows based on differential, projective and Riemaniann geometry developed for active level-set based image segmentation and for 3D reconstruction will first be presented. Applied to brain imaging, these powerful mathematical techniques are then shown to open the possibility of recovering a detailed geometric description of the network of neural connections of the human brain, called Brain connect. I then conclude by positioning the burgeoning field of brain connectomics as an exciting research domain where an open challenge is to build a comprehensive AI based framework to gain insight into brain architecture, functioning and neurodegenerative diseases.

Biography

Dr. Rachid Deriche is Research Director, exceptional class, at Inria, the French National Institute for Research in Digital Science and Technology, where he leads the team Athena dedicated to Computational Brain Imaging. Graduated from Ecole Nationale Supérieure des Télécommunications, Paris, in 1979, he received his Ph.D degree in Mathematics from the University of Paris IX Dauphine, in 1982 and his HDR (Habilitation à Diriger des Recherches) in 1991, from Nice Sophia-Antipolis University (FR).

His expertise and research activities include the domains of Computational Image Processing (CIP), 3D Computer Vision (3D-CV) and Computational Brain Imaging (CBI). His research activities in CIP and 3D-CV have been performed with a particular emphasis in early vision and on the Geometry of multiple views for 3D reconstruction. He successfully advocated the use of differential and projective geometries, variational principles, PDE's and level-sets to solve some of the most CIP and 3D-CV challenging problems. Most of the mathematical concepts and tools he introduced for level-set based image segmentation, PDE based regularization, stereovision, 3D reconstruction, auto-calibration and motion analysis are highly referenced and used world-wide in laboratories and industry. In the early 2000's, he shifted his research interest to computational neuroimaging, focusing on brain imaging, where he significantly contributed to the use of Riemannian geometry in diffusion MRI, and developed and applied AI & mathematical and computational imaging methods to study brain structure and function connectivity.

He has published over 90 journals and 300 peer reviewed conferences papers with a Google Scholar H-index of 79 and nearly 30.000 citations. He has been awarded the 2013 French Academy of Sciences Grand Prize of the EADS Corporate Foundation in Computer Science. In 2014, he has been awarded the Doctorate Honoris Causa by Sherbrooke University (CA). In 2016, he has been awarded the prestigious European Research Council ERC Advanced Grant and in 2019, he has been awarded with a 3IA Senior Chair in AI at the Interdisciplinary Institute for Artificial Intelligence 3IA Côte d'Azur (FR), and the European Association for Signal Processing (EURASIP) elevated him to EURASIP Fellow.

His current research activities are mainly supported within the frameworks of his ERC AdG Grant on Computational Brain Connectivity Mapping and his 3IA Senior Chair on AI based Brain Connectomics.

Title: Building towards a Responsible Data Economy

Abstract

Data is a key driver of modern economy and AI/machine learning, however, a lot of this data is sensitive and handling the sensitive data has caused unprecedented challenges for both individuals and businesses. These challenges will only get more severe as we move forward in the digital era. In this talk, I will talk about technologies needed for responsible data use including secure computing, differential privacy, federated learning, as well as blockchain technologies for data rights, and how to combine privacy computing technologies and blockchain to building a platform for a responsible data economy, to enable more responsible use of data that maximizes social welfare & economic efficiency while protecting users’ data rights and enable fair distribution of value created from data.

Biography

Dawn Song is a Professor in the Department of Electrical Engineering and Computer Science at UC Berkeley. Her research interest lies in AI and deep learning, security and privacy, and blockchain. She is the recipient of various awards including the MacArthur Fellowship, the Guggenheim Fellowship, the NSF CAREER Award, the Alfred P. Sloan Research Fellowship, the MIT Technology Review TR-35 Award, ACM SIGSAC Outstanding Innovation Award, and numerous Test-of-Time Awards and Best Paper Awards from top conferences in Computer Security and Deep Learning. She is an ACM Fellow and an IEEE Fellow. She is ranked the most cited scholar in computer security (AMiner Award). She obtained her Ph.D. degree from UC Berkeley. She is also a serial entrepreneur and has been named on the Female Founder 100 List by Inc. and Wired25 List of Innovators.

Title: Toposes and AI

Abstarct

We shall explain how the very abstract mathematical theory of Grothendieck Toposes arose the interest of some researchers working on AI, and we shall present some aspects of this theory which naturally relate with some fundamental problems of AI.

Biography

Laurent Lafforgue is a French mathematician. He was born in 1966, graduated from Ecole Normale Supérieure in 1990 and got his PhD from Paris XI University in 1993. He was a researcher at CNRS from 1990 to 2000 and a professor at IHES from 2000 to 2021. He was awarded the Fields Medal at the 2002 International Congress of Mathematicians for his work on the “Langlands programme” which relates Galois theory, algebraic geometry and harmonic analysis. He has moved to the Huawei France Fundamental Research Center in Boulogne in Fall 2021.

Jacques N. Biot has international professional experience in higher education and research, life sciences, industry and technology financing, and public administration. The motto of his career has been about how to turn scientific innovation into societal and economic value. He serves as a Board-member or advisor to companies in the field of digital transformation and artificial intelligence, and a Trustee to several scientific academic institutions. He currently chairs the Board of Directors of HuaweiTechnologies France.

Jacques served (2013-2018) as the first executive President of Ecole polytechnique in Paris, the leading public university of science and technology in France, ranked 3rd worldwide by the Times Higher Education for the education of CEOs of Fortune 500 companies. During his tenure he undertook a profound transformation of the institution to make it more science-driven, more entrepreneurial and more international. He developed a strong network of strategic partnerships with leading universities in the US, in Asia, in Israel, in Europe, in Brazil, in Russia, and in Lebanon, and rolled-out a specific collaborative program to support the upgrade of selected establishments in Africa.

He is a member of International Board of the Weizmann Institute in Rehovot and a member of the international Board of the Technion Institute in Haifa, Israel.

From 1992 till 2012, he was the founder, chairman & CEO of JNBD, a strategic consulting firm in life sciences and health technology with a focus on market access issues, which he successfully divested to MAPI, now part of the ICON group. Concurrently, from 2002-2013, he sat as an independent director, chairman of the audit committee and vice chairman of Guerbet Laboratories (GBT, Euronext), a worldwide leader in contrast agents for medical imaging.

He had previously served (1985-1992) as a member of the executive committee in highly innovative international biopharmaceutical companies (Roussel Uclaf and Pasteur Mérieux Serums and Vaccines, both later integrated into the Sanofi Group).

His early career (1977-1985) started within the French administration, in charge of industry and technology funding and company turn-around (CIASI/CIRI) and then as a policy adviser in the office of the Minister of Industry and Research, then in the Prime Minister's office (Laurent Fabius, now chairman of the French Constitutional Court), in charge of industry and technology.

He is the author of numerous reports to the French government on the health product industry.

Jacques is a graduate of Ecole polytechnique (X71) and a member of the elite Corps des Mines.

He is a father of three daughters. His recreational interests are in Performing arts, mountaineering, sailing, and flying. He holds a French & Australian license as a private pilot.

Jacques is an Officer of the Legion of Honor and of the French Order of Merite, a member of the Lion's Order of Senegal and a Commander in the Order of National Education in the Ivory Coast.

Title: Toposes and AI

Abstarct

We shall explain how the very abstract mathematical theory of Grothendieck Toposes arose the interest of some researchers working on AI, and we shall present some aspects of this theory which naturally relate with some fundamental problems of AI.

Biography

Laurent Lafforgue is a French mathematician. He was born in 1966, graduated from Ecole Normale Supérieure in 1990 and got his PhD from Paris XI University in 1993. He was a researcher at CNRS from 1990 to 2000 and a professor at IHES from 2000 to 2021. He was awarded the Fields Medal at the 2002 International Congress of Mathematicians for his work on the “Langlands programme” which relates Galois theory, algebraic geometry and harmonic analysis. He has moved to the Huawei France Fundamental Research Center in Boulogne in Fall 2021.

Title: Toward AI on networks, beyond federated learning

Abstract

Until recently, most of the major advances in machine learning and decision making have focused on a centralized paradigm in which data are aggregated at a central location to train models and/or decide on actions. This paradigm faces serious flaws in many real-world cases. Centralized learning risks exposing user privacy, makes inefficient use of communication resources, creates data processing bottlenecks, and may lead to concentration of economic and political power. It thus appears most timely to develop the theory and practice of a new form of machine learning that targets heterogeneous, massively decentralized networks, involving self-interested agents who expect to receive value (or rewards, incentive) for their participation in data exchanges. In this talk, we will outline the first steps to set the statistical and algorithmic foundations for systems involving multiple incentive-driven learning and decision-making agents, including uncertainty quantification at the agent's level. We will shed new light on the value and handling of data in a competitive, potentially antagonistic multi-agent environment, and sketch some possible directions of research to address these pressing challenges. We will emphasize the interaction of learning with market constraints (scarcity, fairness), connecting adaptive microeconomics and market-aware machine learning. Intelligence on networks build on a decade of joint advances in stochastic optimization, probabilistic machine learning, statistical inference, Bayesian assessment of uncertainty, computation, game theory, and information science. Networked intelligence requires a fundamental departure from standard approaches and leads to major scientific interdisciplinary endeavors that will transform statistical learning in the long term while opening exciting and novel areas of research.

Biography

Eric Moulines earned a degree in engineering (1984) from the Ecole Polytechnique, Paris, and a doctorate in electrical engineering (1990) from the Ecole Nationale Supérieure des Télécommunication. In 1990, he joined the Signal and Image Processing Department at Télécom ParisTech, where he was appointed full professor in 1996. In 2015, he joined the Centre for Applied Mathematics at Ecole Polytechnique, where he is currently Professor of Statistics.

His areas of expertise include computational statistics (Monte Carlo simulations, stochastic optimization), probabilistic machine learning, statistical signal processing and time series analysis (sequential Monte Carlo, non-linear filtering).

His current research topics include high-dimensional Monte Carlo sampling, stochastic optimization, and generative models (variational autoencoders, generative adversarial networks). He applies these different methods to uncertainty quantification, Bayesian inverse problems, and the control of complex systems.

He has published more than 120 articles in leading journals in the areas of signal processing, computational statistics and applied probability and more than 300 proceedings papers in major signal processing and machine learning conferences. In 1997 and 2006, he received the Best paper Award from the IEEE Signal Processing Society (for publications in IEEE Trans. On Signal Processing). He has served on the editorial boards of IEEE Trans. On Signal Processing, Signal Processing, Stochastic Processes and Applications, Journal of Statistical Planning and Inference, Electronic Journal of Statistics. From 2013-2016, he was the Editor-in-Chief of Bernoulli.

E. Moulines is a EURASIP and IMS Fellow. He was awarded the Silver Medal of the Centre National de Recherche Scientifique in 2010 and the Orange Prize of the French Academy of Sciences in 2011. In 2016, he was a Fellow of the IMS. In 2020, he received the technical achievement award from EURASIP. In 2017, he was elected to the Academy of Sciences.