WORKSHOP ON ADVANCED INTELLIGENT SYSTEMS
February 20, 2020
9:00 Opening. Ildar Batyrshin, Program Co-Chair of AIS Workshop
- M. en C. Andrés Ortigoza Campos, Director ESCOM-IPN
- Dr. Marco Antonio Moreno Ibarra, Director CIC-IPN
- Dr. Imre J. Rudas, President of IEEE SMCS
- Dr. Juan Silvestre Aranda Barradas, Secretario de Investigación y Posgrado, IPN
Morning Session 1
Chair: Ildar Batyrshin, CIC IPN, Mexico
9:30-10:15 C.L. Philip Chen, The University of Macau, Macau SAR, China
Unmanned Swarm Planning and Control based on Bionic Movement
C. L. Philip Chen
Member of Academia Europea (AE)
European Academy of Sciences and Arts (EASA)
FIEEE, FAAAS, FIAPR, FCAA, FHKIE
This talk presents swarm control for self-organized system with fixed and switching topologies. The generation strategy, motion control law of a novel leader-follower relation-invariable persistent formation (RIPF), which is a kind of distance-based directed formation for multi-agent systems (MASs), will be discussed. And this formation is similar to the formation of bionic movement. An efficient algorithm is designed to find out if a persistent formation can be generated from a rigid graph. Derived from the properties of a rigid graph, an algorithm to generate a RIPF from any initial location is presented. The communication topology is automatically generated based on RIPF. With the selected minimum agent-movement RIPF, lastly, a control law is designed to drive this initial RIPF to the desired RIPF with given distances among agents. Simulation results show the proposed generative method, control law, and downward-tree are effective to realize the desired formation.
10:15-11:00 Robert Kozma, Department of Computer Science, UMass Amherst, MA, USA
Intelligent Systems with Flexible Adaptation in Response to Rapidly Changing Environmental Conditions in Emergency Scenarios
Dr. Robert Kozma
Fellow of IEEE, Fellow of the International Neural Networks Society, INNS
Dr. Robert Kozma holds PhD in Physics (Delft, The Netherlands, 1992), MSc Mathematics (Mathematics, Budapest, Hungary, 1988; Power Engineering, Moscow, Russia, 1982). He is Professor of Computer Science, University of Massachusetts Amherst, where he is Director of the Biologically-Inspired Biological and Dynamical Systems (BINDS) Lab since 2016. His previous faculty positions include University of Memphis, TN; University of California at Berkeley, CA; Otago University, Dunedin, New Zealand; Tohoku University, Sendai, Japan. He has held visiting positions at NASA/JPL Robotics, Lawrence Berkeley Laboratory, Air Force Research Lab/WPAFB, and Sarnoff Co., Princeton, NJ, US. His research focuses on the design, analysis, and control of artificial and biological intelligent systems, as well as robust decision support systems and networks. He is author of 3 patent submissions, 7 book volumes, over 300 publications in journals and refereed proceedings. He is Editor-in-Chief of IEEE Transactions on Systems, Man, and Cybernetics: Systems (2020-2022); served on the SMC Board (2016-2018), on the BOG of INNS (2007-2012), CIS AdCom (2009- 2012); Past-President of INNS (2017-2018). He received the INNS Dennis Gabor Award (2013).
Cutting-edge intelligent systems using advanced AI exhibit outstanding performance in many important tasks when the conditions are well-defined. However, the performances of these intelligent systems may drastically deteriorate when the data are perturbed, or the task/environment dynamically changes, either due to natural effects including noise, or due man-made disturbances. We address the issue of dynamic adaptation and flexibility of intelligent systems, in order to preserve their integrity, maintain their operation, and provide robust response to unexpected, often unpredictable events. The proposed approach views intelligent systems as large-scale networks and applies percolation theory and phase transitions over random graphs to describe the spatio-temporal processes in the system. The developed models are used to characterize, potentially predict, and mitigate the consequences of rare events with sudden, drastic changes in the behavior of complex systems. Implementations include distributed, multi-modal sensor systems, and the development of decision support systems. Broader applications include the analysis and prediction of stock market, solar flares, earthquakes, and supernova dynamics in distant Galaxies.
11:00-11:30 Coffee break
Morning Session 2
Chair: Felix Castro, Pachuca Autonomous University of Hidalgo, Pachuca, Mexico
11:30-12:15 Vladik Kreinovich, University of Texas at El Paso, USA
Deep Learning (Partly) Demystified
Vladik Kreinovich is a Professor of Computer Science at the
University of Texas at El Paso. His main interests are the representation and processing of uncertainty, especially interval computations and intelligent control. He has published eight books, 24 edited books, and more than 1,500 papers. Vladik is the Vice President of the International Fuzzy Systems
Association (IFSA), Vice President of the European Society for
Fuzzy Logic and Technology (EUSFLAT), Fellow of International Fuzzy
Systems Association (IFSA), Fellow of Mexican Society for
Artificial Intelligence (SMIA), Fellow of the Russian Association
for Fuzzy Systems and Soft Computing. He is Treasurer of
IEEE Systems, Man, and Cybernetics Society.
Successes of deep learning are partly due to appropriate selection of activation function, pooling functions, etc. Most of these choices have been made based on empirical comparison and heuristic ideas. In this paper, we show that many of these choices — and the surprising success of deep learning in the first place — can be explained by reasonably simple and natural mathematics.
12:15-13:00 Ildar Batyrshin, CIC-IPN, Mexico
Problem Dependent and Data-Driven Similarity and Correlation Measures in Knowledge Retrieval
Afternoon Session 1
Chair: Alexander Gelbukh, CIC IPN, Mexico
14:00-14:45 Alex Poznyak, CINVESTAV-IPN, Mexico
“Is It Possible to Control Dynamic Systems without Exact Knowledge and Designing of Mathematical Models? Yes, it’s possible: Sliding Mode Approach and Differential Neural Networks”
Prof. Alex S. Poznyak
Dept. Control Automatico
CINVESTAV-IPN, AP-14-740, Mexico D.F. Mexico
Tel: (52-5) 747-37-36, 747-37-36
FAX: (52-5) 747-70-89
Alexander S. Poznyak graduated from Moscow Physical Technical Institute (MPhTI) in 1970. He earned Ph.D. and Doctor Degrees from the Institute of Control Sciences of Russian Academy of Sciences in 1978 and 1989, respectively. From 1973 up to 1993 he served this institute as researcher and leading researcher, and in 1993 he accepted a post of full professor (3-F) at CINVESTAV of IPN in Mexico. 8years he was the head of the Automatic Control Department. He is the director of 42 PhD thesis’s (39 in Mexico). He has published more than 240 papers in different international journals and 14 books in leading Publisher Companies such as Springer, Elsevier, Birkhauser, and Marcel Decker. Alexander is a Regular Member of Mexican Academy of Sciences and System of National Investigators (SNI-Emerito from 2014). Prof. Poznyak is a Fellow of IMA (Institute of Mathematics and Its Applications, Essex UK) and Associated Editor of Oxford-IMA Journal on Mathematical Control and Information, of Kybernetika (Chech Republic), Nonlinear Analysis: Hybrid systems (IFAC) as well as Iberamerican Int. Journal on “Computations and Systems”. He was also Associated Editor of CDC, ACC and Member of Editorial Board of IEEE CSS. He is a member of the Evaluation Committee of SNI (Ministry of Science and Technology, area 7 and Emeritus Committee) responsible for Engineering Science and Technology Foundation in Mexico, and a member of Award Committee of Premium of Mexico on Science and Technology. In 2014 he was invited by the USA Government to serve as the member of NSF committee on “Neuro Sciences and Artificial Intelligence”.
The possibility of creating control systems without first developing a
mathematical model (and as a result, without identifying its parameters) is one of the key approaches to creating Artificial Intelligence Systems with direct application of Digital Technologies. Among the few approaches to the creation of such systems, a special place, due to its simplicity and as a result of its popularity, take Sliding Modes Method (Sliding Mode – SM Approach) and Dynamic Neural Networks (DNN Approach) which will be discussed in the talk. Both approaches require minimal information about a dynamic system for their implementation: the number of states, the number of controls and outputs available to observation. The SM method is illustrated by the example of a problem of controlling the motion of a rigid body. DNN-Learning is illustrated by the process of photo catalytic ozonization during water purification from organic pollutants.
14:45-15:30 Saeid Nahavandi, Deakin University – Victoria, Australia
Haptically-Enabled VR-Based Simulators
Alfred Deakin Professor
Institute for Intelligent Systems Research and Innovation
Deakin University – Victoria, Australia
Saeid Nahavandi received his BSc (Hons), MSc and PhD in Control Engineering from Durham University, UK in 1985, 1986 and 1991 respectively. Saeid is an Alfred Deakin Professor, Pro Vice-Chancellor and the Director for the Institute for Intelligent Systems Research and Innovation at Deakin University in Australia. Professor Nahavandi is a Fellow member of IEEE, IET and IEAust. He has published over 800 refereed papers and been awarded over 50 competitive grants over the past 30 years. He received the Research collaboration / initiatives award from Japan and Prince & Princess of Wales Science Award and two Life time Achievements Awards. Saeid won the title of Young Engineer of the Year Award in 1996 and holds six patents, two of which have resulted in two very successful start-ups, (Universal Motion Simulator Pty Ltd and FLAIM Systems Pty Ltd).
Saeid has carried out industry based research with several major international companies such as Airbus, Boeing, Bosch, Ford Motor Company, General Motors, General Dynamics, Holden, Lockheed Martin, Nissan, Thales and Vestas, just to name a few. Professor Nahavandi is the General Chair for IEEE SMC 2021. He was also the General Co-Chair for IEEE SMC 2011. He holds the positions of Senior Associate Editor: IEEE Systems Journal, Editor-In-Chief for IEEE SMC Magazine, Chair: IEEE SMC Electronic Communications Subcommittee and founding Chair: IEEE SMC Victorian Chapter. Saeid is a Fellow of the Australian Academy of Technological Sciences and Engineering (FTSE).
This talk will focus on integration of haptic technology into augmented or virtual reality systems to increase their fidelity and realism. The inventions enable users to “touch-and-feel” virtual and remote objects and perceive their attributes, such as texture and hardness,engaging users in an immersive virtual environment. One such application is the use of such technology for fire fighting training applications.
Fire fighting is a physically demanding task that requires extensive training. The current manual training procedures do not take into consideration the immersion factor, without which a novice fire fighter may be overwhelmed when facing a fire for the first time. This challenge was one of the main motivations to harness the power of virtual reality and haptics to develop a portable fire fighting training system. This presentation covers
technologies used in development of a haptically enabled VR based fire fighting training system that enables the trainees to experience the jet reaction forces from the hose and provides realistic water dispersion and interaction with fire and smoke particles via accurate
particle physics modelling.
15:30-16:00 Coffee break
Afternoon Session 2
Chair: Grigori Sidorov, CIC IPN, Mexico
16:00-16:45 Alexander Gelbukh, CIC IPN, Mexico
Advances in Sentiment Analysis and Author Profiling in Social Networks
16:45 – 17:30 Gyorgy Eigner and Levente Kovacs, Obuda University, Hungary
Cyber-Medical Systems: The Digitalized Healthcare Approach and Its Trends
Prof. Dr. habil. Levente Kovács received his M.Sc. degrees in Electrical Engineering in 2000 (“Politehnica” University of Timişoara, Romania) and Biomedical Engineering in 2011 (Budapest University of Technology and Economics, Hungary). He received his PhD from the Budapest University of Technology and Economics in 2008. His fields of interest are modern control theory and physiological controls – within these subjects, he has published more than 400 articles in international journals and refereed international conference papers. Currently, he is the president of Óbuda University. He is full professor of the John von Neumann Faculty of Informatics at Óbuda University. He founded the Physiological Controls Research Center in Óbuda University in 2013 being the head of it and also was János Bolyai Research Fellow of the Hungarian Academy of Sciences. Prof. Kovács is the chair of IEEE Hungary Section and the chair of the IEEE SMC Hungary Chapter. He is 2015 recipient of the highly prestigious ERC StG grant of the European Union.
Dr. György Eigner graduated from Bánki Donát Faculty of Mechanical and Safety Engineering of Óbuda University, and he received his Master Degree in Biomedical Engineering from the Budapest University of Technology and Economics. He got his PhD degree at Doctoral School of Applied Informatics and Applied Mathematics of Óbuda University. He works as researcher at the Physiological Controls Research Center. Dr. Eigner holds an assistant professor position at the John von Neumann Faculty of Informatics of Óbuda University, where he also fills the deputy head of the Biomatics Institution. Besides, he serves as advisor at different industry related firms. During his career has published more than 75 papers in various journals and international conference proceedings. He is an associate editor of the Acta Polytechnica Hungarica and IEEE Access international journal.
By the needs of Industry 4.0 digitalization of industry has been accentuated. However, cybersphysical systems left its footprints in healthcare as well, formalizing the cyber-medical system requirements: how to help everyday life of medical doctors and patients by using medical devices in a more personalized way.
This concept meshes the entire biomedical engineering research field. The idea is to create mathematical algorithms able to be personalized on the patients’ need and physiology, use cloud computation techniques to fasten the decision support and big data analysis for feature extraction. As a result, the cyber-medical system concept is equivalent with a smart healthcare framework whereby using the computational power possibilities together with machine learning, artificial intelligence and control engineering methods we would like to intensify the decision support of doctors, nurses and patients in order to use in an intelligent way the knowledge-based medical applications.
The presentation will give an overview on the above-mentioned aspects: how the cyber-medical system concept appeared, what the trends are, how it is influencing our every day’s life including education, research, medical industry and healthcare sector. In order to give concrete examples, individualized model-based applications are presented such as:
- Artificial pancreas problem: Recent technological advances in diabetes treatment like Continuous
Glucose Monitors (CGMs) for the subcutaneous measurement of glucose concentration and insulin pump for the subcutaneous delivery of insulin allowed investigating the applicability of an external controller. In type 1 diabetes, where the disease can be characterized as a general clinical picture (e.g. complete pancreatic
β-cell insufficiency) different individualized model-based (mostly model predictive control (MPC) based) solutions have been already formulated and even clinical trials appeared demonstrating its applicability. However, closing the loop needs the integration of individualized control methods and robust control algorithms in order to successfully reject hypoglycemic episodes. Moreover, the scientific community started to transpose the problem on type 2 diabetes as well, trying to cover the largest diabetic population with the artificial pancreas concept.
- Tamed cancer concept: Beside general cancer therapies like chemotherapy, targeted molecular therapies appeared in order to focus on a given mechanism of the tumor growth stopping it in a more effective / individual way. Antiangiogenic therapy is one targeted molecular therapy arose in the last decade which aims to stop tumor angiogenesis, i.e. formation of new blood vessels; hence, minimizing the tumor’s size. However, the corresponding drugs are very expensive, and in high doses may have side
effects; moreover, the currently used clinical protocols are determined empirically. Consequently, this
problem can be seen as a double optimal control problem: on the one hand, the aim is to minimize the
tumor’s volume by a model-based control algorithm, but it is also crucial to inject the corresponding inhibitor (drug) in the best way (minimizing costs). Due to the heterogeneous nature of the patients by a recently gained ERC StG grant of the European Union we try to develop a novel concept: taming cancer. Beside creating a more realistic tumor growth mathematical model, we would like to decrease the injection sampling time of inhibitors that combined with a robust control method can control the tumor volume artificially and in this taming it. As a result, it would be even possible living with the tumor in a controlled way.
- Ildar Batyrshin, CIC-IPN, SMIA
- Felix Castro, UAEH, SMIA
Local Organizing Committee Chairs:
- Juan Silvestre Aranda Barradas, IPN
- Marco Antonio Moreno Ibarra, CIC-IPN
- Andrés Ortigoza Campos, ESCOM-IPN
- Abraham Rodríguez Mota, CIC-IPN
Local Organizing Committee Members:
- Elvia Cruz Morales
- Maldonado Sifuentes Christian Efraín
Instituto Politécnico Nacional, Escuela Superior de Computo
Av. Juan de Dios BAtiz s/n esq. Av. Miguel Othon de Mendizabal. Colonia Lindavista. Alcaldia: Gustavo A. Madero. C. P. 07738. Ciudad de México.
Tel. 57296000 Ext. 46188 Email: email@example.com