The Cristoffel-Darboux kernel as a tool in data analysis
Jean-Bernard LASSERRE,
Directeur de Recherche emeritus, LAAS-CNRS
Abstract: If the Christoffel-Darboux (CD) kernel is well-known in theory of approximation and orthogonal polynomials, its striking properties seem to have been largely ignored in the context of data analysis (one main reason being that in data analysis one is faced with measures supported on finitely many points (the data set)). In this talk, we briefly introduce the (CD) kernel, some of its properties, and claim that it can become a simple and easy to use tool in the context of data analysis, e.g. to help solve some problems like outlier detection, manifold learning and density estimation.
Biography: Jean-Bernard Lasserre graduated from "Ecole Nationale Supérieure d'Informatique et Mathematiques Appliquees" (ENSIMAG) in Grenoble, then got my PhD (1978) and "Doctorat d'Etat" (1984) degrees both from Paul Sabatier University in Toulouse (France). Jean-Bernard has been at LAAS-CNRS in Toulouse since 1980, where he is currently Directeur de Recherche (emeritus). He is also a member of IMT, the Institute of Mathematics of Toulouse, and holds the ``Polynomial Optimization" chair at the ANITI Institute (one of the four recently created Artificial Intelligence Institutes in France). Jean-Bernard’s past and present research activities cover machine Learning, applied mathematics, control and non-linear PDEs, probability, Markov control processes, approximation theory & convex optimization, production planning & scheduling and author/co-author of nine books in these area. In particular, he has initiated the ``Moment-SOS hierarchy" also known as "Lasserre hierarchy", a novel methodology used in many areas for solving hard nonconvex polynomial optimization problems. Jean-Bernard was twice a one-year visitor (1978-79 and 1985-86) at the Electrical Engineering Department. of the University of California at Berkeley with a fellowship from INRIA and NSF. He has done several one-month visits to Stanford University (Stanford, California), the Massachusetts Institute of Technology (MIT, Cambridge), the Mathematical Sciences Research Institute (MSRI, Berkeley), the Fields Institute (Fields, Toronto), the Institute for Mathematics and its Applications (IMA, Minneapolis), the Institute for Pure and Applied Mathematics (IPAM, UCLA), Cinvestav-IPN (Cinvestav, Mexico), Leiden University (Leiden, The Netherlands), the Tinbergen Institute (Tinbergen, Amsterdam, The Netherlands), the University of Adelaide (Adelaide, Australia), the University of South Australia (UniSA, Adelaide), the University of New South wales (UNSW, Sydney), the University of British Columbia (UBC, Vancouver).
Awards & Distinctions: • ISSAC' 2019 Distinguished Paper Award (Beijing, July 2019) with Florent Bréhard and Mioara Joldès • Simons CRM Professor at the CRM in Montreal (October 2019) • Invited Speaker at the International Congress of Mathematicians (ICM 2018), Rio de Janeiro, August 2018: Section 16: Control Theory & Optimization. • 2015 John von Neumann Theory prize of the INFORMS society • 2015 Khachiyan prize of the Optimization Society of INFORMS • 2009 Lagrange prize in Continuous optimization (awarded jointly every 3 years by SIAM and the Mathematical Optimization Society) • SIAM Fellow (class 2014) • 2014 Laureate of an ERC-Advanced Grant from the European Research Council (ERC) for the TAMING project.
Mathematical Optimization for a More Transparent Data-Driven Decision-Making
Dolores ROMERO MORALES,
Professor, Copenhagen Business School
Abstract: In this presentation, we will review recent advances from the Mathematical Optimization community to strike a better balance between transparency and accuracy of Data Science models. Despite excellent accuracy, state-of-the-art Data Science models are effectively acting as black boxes. This hinders model validation in Data Driven Decision Making, but it may also transmit biases in the data and give unfair predictions to individuals in risk groups. Therefore, transparency is required by regulators for models aiding, for instance, credit scoring, and since 2018 the EU has extended this requirement by imposing the so-called right-to-explanation in algorithmic decision-making. We will review different interpretations of transparency and how to achieve them with the help of Mathematical Optimization. First, we will provide local explanations to understand how the Data Science model arrived at individual predictions. Second, we will obtain counterfactual explanations to give to an individual or a group of individuals feedback on how to get with the same model another prediction by making the changes to the features of the individuals. Third, we will model cost-sensitive and fairness constraints, to avoid the predictions discriminate against risk groups.
Biography: Dolores Romero Morales is a Professor in Operations Research at Copenhagen Business School. Her areas of expertise include Supply Chain Optimization, Data Mining and Revenue Management. In Supply Chain Optimization she works on environmental issues and robustness. In Data Mining she investigates interpretability and visualization. In Revenue Management she works on large-scale network models. Her work has appeared in a variety of leading scholarly journals, including European Journal of Operational Research, Management Science, Mathematical Programming and Operations Research, and has received various distinctions. Currently, she is Editor-in-Chief to TOP, the Operations Research journal of the Spanish Society of Statistics and Operations Research, and an Associate Editor of Omega.
She has worked with and advised various companies on these topics, including IBM, SAS, KLM and Radisson Edwardian Hotels, as a result of which these companies managed to improve some of their practices. SAS named her an Honorary SAS Fellow and member of the SAS Academic Advisory Board. She currently leads the EU H2020-MSCA-RISE NeEDS project, which has a total of 14 participants and a budget of more than €1.000.000 for intersectoral and international mobility, with the aim to improve the state of the art in Data Driven Decision Making.
Dolores joined Copenhagen Business School in 2014. Prior to coming to Copenhagen Business School, she was a Full Professor at University of Oxford (2003-2014) and an Assistant Professor at Maastricht University (2000-2003). She has a BSc and an MSc in Mathematics from Universidad de Sevilla and a PhD in Operations Research from Erasmus University Rotterdam.
Quality Science for Smart Manufacturing in the Era of Data-Driven Automation
Jianjun SHI,
The Carolyn J. Stewart Chair and Professor
A Member of the National Academy of Engineering of USA
Georgia Institute of Technology
Abstract: In smart manufacturing systems, a large number of different types of sensors are deployed for monitoring machine status, process variables, product quality, and the overall system performance. Analyzing for effective cost reduction and quality improvement with those massive amounts of heterogeneous data, arriving rapidly and manifesting with spatio-temporal complexity, is undoubtedly challenging in all manufacturing enterprises. This presentation will discuss research opportunities, challenges, and advancements in this important research area of quality science, but more so in the context of how machine learning and data-driven automation have reshaped the research landscape. Examples of ongoing research projects will be used to illustrate and exemplify the frontiers of this research area. All examples come from real data and real-life industrial production systems. This presentation will emphasize what motivates the research undertakings as well as why the research matters, including the challenges to be overcome, new ideas behind the methods developed, and validation/implementation undertook.
Biography: Dr. Jianjun "Jan" Shi is the Carolyn J. Stewart Chair and Professor in the H. Milton Stewart School of Industrial and Systems Engineering, with a joint appointment in the George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology. Prior to joining Georgia Tech in 2008, he was the G. Lawton and Louise G. Johnson Professor of Engineering at the University of Michigan. He received his B.S. and M.S. in Automation from the Beijing Institute of Technology in 1984 and 1987, and his Ph.D. in Mechanical Engineering from the University of Michigan in 1992. Professor Shi’s research is in the development and application of data-enabled manufacturing. His methodologies integrate system informatics, advanced statistics, and control theory for the design and operational improvements of manufacturing and service systems by fusing engineering systems models with data science methods. The technologies developed by Dr. Shi’s research group have been widely implemented in various production systems with significant economic impacts.
Professor Shi is the founding chairperson of the Quality, Statistics and Reliability (QSR) Subdivision at the Institute for Operations Research and Management Science (INFORMS). He is currently serving as the Editor-in-Chief of IISE Transactions, the flagship journal of the Institute of Industrial and Systems Engineers. He is a Fellow of four societies: IISE, ASME, ISI, and INFORMS, and an Academician of the International Academy for Quality, and a member of the National Academy of Engineering (NAE) of USA. Dr. Shi has received various awards for his research and teaching, including the ASQ Brumbaugh Award (2019), the Horace Pops Medal Award (2018), the IISE David F. Baker Distinguished Research Award (2016), the IIE Albert G. Holzman Distinguished Educator Award (2011), the Forging Achievement Award from Forging Industry Educational and Research Foundation (2007), Monroe-Brown Foundation Research Excellence Award (2007), the 1938E Award (1998) at The University of Michigan, and NSF CAREER Award (1996).
Challenges and potentials of data-driven automation in complex manufacturing systems
Philippe VIALLETELLE,
STMicroelectronics
Abstract:
Among the most complex industries existing today, major players in the semiconductor industry learned early on that only by working together could they succeed in integrating hundreds of machines together with automated transportation, real-time management of their production flows as well as advanced process control techniques relying on data coming from thousands of sensors throughout the various systems. Today it is estimated that more than 80% of all the data collected is never really used or even looked at. One of the main causes may be that the first manufacturing systems were siloed systems. Interpreting data and navigating these layers and information silos required deep technical expertise and considerably limited the development of wide-scale analytics. How human knowledge and engineering know-how is integrated into these systems is seen as one of the next challenges for the semiconductor industry. Exploring how new technologies can help us benefit from data-driven automation and look closer at the fantastic promises of Artificial Intelligence, Automation Science, Knowledge Management and Operations Research is at the core of this presentation.
Biography: Philippe VIALLETELLE is a Senior Member of Technical Staff at STMicroelectronics Crolles, France. He received an Engineering degree in Physics from the INSA Rennes in 1989 and spent his entire career in the semiconductor industry where he occupied various positions, mostly in manufacturing support functions. Since 2005 he has been defining and driving research and innovation activities for STMicroelectronics in various projects at European level including the modelling and optimization of manufacturing systems such as factory automation and industry 4.0. At STMicroelectronics, he is now actively contributing to the development of a global semantic model within the Manufacturing Data & Analytics program which is today one of the key strategic priorities of the company.
