DEA under Big Data: Data Enabled Analytics and Network Data Envelopment Analysis
Prof. Joe Zhu
Worcester Polytechnic Institute, USA
This talk proposes that data envelopment analysis (DEA) should be viewed as a method (or tool) for data-oriented analytics in performance evaluation and benchmarking. While computational algorithms have been developed to deal with large volume of data (decision making units, inputs, and outputs) under the conventional DEA, valuable information hidden in big data that are represented by network structures should be extracted by DEA. These network structures, e.g., transportation and logistics systems, encompass a broader range of inter-linked metrics that cannot be modelled by the conventional DEA. It is proposed that network DEA is related to the value dimension of big data. It is shown that network DEA is different from the standard DEA, although it bears the name of DEA and some similarity with the conventional DEA. Network DEA is big Data Enabled Analytics (big DEA) when multiple (performance) metrics or attributes are linked through network structures. These network structures are too large or complex to be dealt with by the conventional DEA. Unlike the conventional DEA that are solved via linear programming, general network DEA corresponds to nonconvex optimization problems. This represents opportunities for developing techniques for solving non-linear network DEA models. Areas such as transportation and logistics system as well as supply chains have a great potential to use network DEA in big data modeling.
Biography of Prof. Joe Zhu
Joe Zhu is Professor of Operations Analytics in the Foisie Business School, Worcester Polytechnic Institute. He is the founding Dean of College of Auditing and Evaluation, Nanjing Audit University. He is an internationally recognized expert in methods of performance evaluation and benchmarking using Data Envelopment Analysis (DEA). With more than 30,000 Google Scholar citations, he is recognized as one of the top authors in DEA with respect to research productivity, h-index, and g-index. In 2017, he is ranked No. 3 among the most productive and influential authors in 40 years of European Journal of Operational Research. His research has been supported by KPMG Foundation, National Institute of Health, and Department of Veterans Affairs. He has extensive editorial experience, as the Deputy Editor of OMEGA, Editorial Board Member of European Journal of Operational Research, and Computers and Operations Research, Associate Editor of Journal of the Operational Research Society, and INFOR, Series Associate Editor of Springer's International Series in Operations Research and Management Science, and others. He has published over 140 peer-reviewed articles in Operations Research, European Journal of Operational Research, Journal of the Operational Research Society, IIE Transactions, Management Science, Sloan Management Review, Annals of Operations Research, OMEGA, International Journal of Production Economics, Naval Research Logistics, and others. He is a Japan Society for Promotion of Science (JSPS) fellow, a William Evans Visiting Fellow (University of Otago, New Zealand), Feng Tay Chair Professor (National Yunlin University of Technology and Science, Taiwan), and Chang Jiang Scholar Chair Professor awarded by the Ministry of Education of China.
Modern Sliding Mode Control with Application to Connected and Automated Vehicles
Prof. Antonella Ferrara
University of Pavia, Italy
Sliding Mode Control is a nonlinear control methodology based on the use of a discontinuous control input which forces the controlled system to switch from one structure to another, evolving as a variable structure system. This structure variation makes the system state reach in a finite time a pre-specified subspace of the system state space where the desired dynamical properties are assigned to the controlled system. In the past years, an extensive literature has been devoted to the developments of Sliding Mode Control theory. This kind of methodology offers a number of benefits, the major of which is its robustness versus a significant class of uncertaintiesduring the sliding mode. Yet, it presents a crucial drawback, the so-called chattering, which may disrupt or damage the actuators and induce unacceptable vibrations throughout the controlled system. This limits the practical applicability of the methodology, especially in case of mechanical or electromechanical plants. This drawback has been faced by the theoretical developments of the last two decades. They will be reviewed in this talk, illustrating how the “modern” results can beprofitably used to solve practical control problems in complex systems as the automotive ones. The talk will develop starting from single vehicle control and coming to discuss formation control of fleets of connected and automated vehicles (CAVs), with the aim of producing a beneficial impact on road traffic..
Biography of Prof. Antonella Ferrara
Antonella Ferrara received the M.Sc. in Electronic Engineering (cum Laude) in 1987 and the Ph.D. in Computer Science and Electronics in 1992 from the University of Genova. After being a Faculty Member at the University of Genova (1992-1998) shejoined the University of Pavia in 1998 as an Associate Professor. Since January 2005 she has been Full Professor of Automatic Control in the Department of Electrical, Computer and Biomedical Engineering (ECBE) of the same university. She teaches "Process Control and Robotics" and "Advanced Automation and Control". She was Visiting Professor at Graz University of Technology, and Invited Lecturer at Harvard University, University of Minnesota, University of California at Los Angeles (UCLA), University of Stuttgart, Technical University of Delft, INRIA Grenoble, King Abdullah University of Science and Technology (KAUST), Jeddah, Hanyang University, Seoul, and Universidade Federal do Rio de Janeiro, Brazil. Her research activities are mainly in the area of automatic control of complex systems, with application to automotive systems, robotics, power networks and vehicular traffic. She is particularly interested in autonomous driving and coordinated control of vehicles. She was coordinator of the Italian research team in the European Projects ITEAM (Interdisciplinary Training Network in Multi-Actuated Ground Vehicles, Marie Skłodowska-Curie Action, 2016-2019). She was scientific leader of several research projects funded or co-funded by companies. She has authored or co-authored more than 380 scientific papers, with 125 international journal papers and 3 books. Moreover, she contributed, with invited chapters, to several edited books. She was Associate Editor of the IEEE Transactions on Control Systems Technology, IEEE Transactions on Automatic Control and IEEE Control Systems Magazine.She is currently Associate Editor of Automatica and IEEE Transactions on Intelligent Vehicles, as well asSubject Editor of the International Journal of Robust and Nonlinear Control. She wasappointed and elected member of the Board of Governors of the IEEE Control Systems Society (2012; 2016-2018), and member of the European Control Association (EUCA) Council (2015-2019). She is Chair of the EUCA Conference Editorial Board and Series Editor of the book series "Advances in Industrial Control" published by Springer.
Deepening Smart Data: Quality data to enhancing Deep Learning applications
Prof. Francisco Herrera
University of Granada, Spain
In the last years, deep learning methods and particularly Convolutional Neural Networks (CNNs) have exhibited excellent accuracies in many image and pattern classification problems, among others.
The strengthening smart data to get quality datais the foundation for good data analytics in general. It is very important for getting a good deep learning model from an applied point of view. Consolidating smart data requires a data preprocessing analysis to adapt the raw data to fulfill the input demands of each learning algorithm.
In contrast to the classical classification models, the high abstraction capacity of CNNs allows them to work on the original high dimensional space, which reduces the need for manually preparing the input. However, a suitable data preprocessing is still important to improve the quality of the model.One of the most used preprocessing techniques with CNNs is data augmentation for small image data sets, which increases the volume of the training data set by applying several transformations to the original input. There are other data guided preprocessing procedures for specific problems, like image brightness or contrast, noise adaptation, among others.
In this talk we present the connection between deepening smart data and heightening deep learning, focused on some applications enhancing. Among them, we will present some video surveillance applications and the best result in the MNIST problem with only 10 errors.
Biography of Prof. Francisco Herrera
Francisco Herrera received his M.Sc. in Mathematics in 1988 and Ph.D. in Mathematics in 1991, both from the University of Granada, Spain. He is currently a Professor in the Department of Computer Science and Artificial Intelligence at the University of Granada and Director of DaSCI Institute (Andalusian Research Institute in Data Science and Computational Intelligence). He is an EurAI Fellow 2009 and IFSA Fellow 2013. He's an academician at the Spanish Royal Academy of Engineering (May 2019).
He has been the supervisor of 50 Ph.D. students. He has published more than 500 journal papers that have received more than 80000 citations (Scholar Google, H-index 139). He is coauthor of the books "Genetic Fuzzy Systems" (World Scientific, 2001) and "Data Preprocessing in Data Mining" (Springer, 2015), "The 2-tuple Linguistic Model. Computing with Words in Decision Making" (Springer, 2015), "Multilabel Classification. Problem analysis, metrics and techniques" (Springer, 2016), "Multiple Instance Learning. Foundations and Algorithms" (Springer, 2016), “Learning from Imbalanced Data Sets” (Springer, 2018) and “Big Data Preprocessing – Enabling Smart Data” (Springer, 2020).
He currently acts as Editor in Chief of the international journal "Information Fusion" (Elsevier). He acts as editorial member of a dozen of journals.
He has been selected as a Highly Cited Researcher (in the fields of Computer Science and Engineering, respectively, 2014 to present, Clarivate Analytics).
His current research interests include among others, computational intelligence (including fuzzy modeling, evolutionary algorithms and deep learning), computing with words, information fusion and decision making, and data science (including data preprocessing, prediction and big data).
Privacy and safety in cyber-physical systems
Prof. Dimitri Lefebvre
University of Le Havre Normandie, France
In recent years, the tremendous growth of computer technology has led to the proliferation of highly complex cyber-physical systems, in particular in Industry 4.0 and Internet of Things. Such systems exhibit behaviors determined by the asynchronous occurrence of certain events and are termed Discrete Event Systems(DES). Examples of DES are encountered in many traditional application domains, such as automated manufacturing, computer networks, transportation, as well as in emerging areas like healthcare, or management of technical, human and financial resources.
A significant research effort has been devoted to DES in order to address a series of difficult problems that are often combinatorial in nature, and require advances methodologies The aim of this talk is first to give some basic modeling notions to represent DES in order to address a large variety of problems for cyber-physical systems. Then, some privacy,and safety issues are discussed. On one hand, opacity is presented to increase privacy and cyber-security. On the other hand, fault diagnosis illustrates how it is possible to detect and isolate some faults in the system based on its observations. Performance evaluation is introduced to provide a detailed analysis of such properties.
Finally, a list of open questions and future challenges is proposed.
Biography of Prof. Dimitri Lefebvre
Dimitri Lefebvre received the S.B. in Science and Engineering in 1990, the M.Eng. degree in Automatic Control and Computer Science in 1992, and the Ph.D. degree Automatic Control and Computer Science in 1994, all from University of Sciences and Technologies and Ecole Centrale in Lille, France. In 1995, he joined the University of FrancheComté, Belfort, France, where he served as Associate Professor with the Department of Electrical Engineering and the Research Group about Systems and Transportations. Since 2001, he has been with University Le HavreNormandie, France as Professor. He is currently with the Research Group on Electrical Engineering and Automatic Control (GREAH) in Le Havre and was from 2007 to 2012 the head of the group. His current research interests include fault diagnosis and control designfor dynamic systems, discrete event systems, learning processes and artificial intelligence, with applications to network security and safety in the domains ofelectrical engineering, robotics, transportations and logistics. He is the authors of more than 100 articles published in indexed journals and more than 200 communications in international conferences.