Home Invited lectures

IEEE-CEC 2012 Invited Lectures

 Exploring the Issue of Representation in Evolutionary Computation

Dan Ashlock

Abstract: The choice of representation and accompanying variation operations defines the fitness landscape and search space of an evolutionary computation problem.  Careful design of a representation permits the researcher to

* Embed domain knowledge into the system.
* Control the size of the search space.
* Enhance the evolvability of entities in the space.
* Influence the topology of the fitness landscape.

Accessible examples of representations for problems in game theory, real optimization, automatic content generation for games, genetic programming, and bioinformatics will be given.  Different representations will be compared and contrasted and design principles for representation will be discussed.

This presentation includes discussions of generative representations, fractal representations, and non-standard representations for domains such as genetic programming.  Examples of situations in which the choice of representation dominates the behavior of an evolutionary computation system are given.  Multiple representations for the same problem will be contrasted to demonstrate the impact and value of testing different representations.

Presenter Biography:
Daniel Ashlock is a Professor of Mathematics and holds the Bioinformatics Chair in the Department of Mathematics and Statistics at the University of Guelph in Canada.  He holds a doctorate from Caltech in pure mathematics.  He is the author of over 175 peer-reviewed articles in mathematics, evolutionary computation, bioinformatics, and theoretical biology.  He currently serves as an associate editor of the IEEE Transactions on Evolutionary Computation, the IEEE Transactions on Computational Intelligence and Artificial Intelligence in Games, The IEEE/ACM Transactions on Computational Biology and Bioinformatics, and the Journal Biosystems.  His research interests include representation in evolutionary computation, spatially structured algorithms, bioinformatics, ecoinformatics, and graph theory.  Most of his research is structured as collaborations where new ideas flow from helping other people solve their problems.

Some thoughts on a gap between theory and practice of evolutionary algorithms

Zbyszek Michalewicz

Abstract: At the Workshop on Evolutionary Algorithms, organized by the Institute for Mathematics and Its Applications, University of Minnesota, Minneapolis, Minnesota, October 21 – 25, 1996, one of the invited speakers, Dave Davis made an interesting claim. As the most recognised practitioner of evolutionary algorithms at that time he said that all theoretical results in the area of evolutionary algorithms were of no use to him – actually, his claim was a bit stronger. He said that there was one use only of such results: if a theoretical result indicated that, say, the best value of some parameter was such-and-such, he would never use the recommended value in any real-world implementation of evolutionary algorithm! Clearly, there was – in his opinion – a significant gap between theory and practice of evolutionary algorithms.

After more than 15 years from that time it is worthwhile to revisit this claim and to answer some questions; these include: What are the practical contributions coming from the theory of evolutionary algorithms? Did we manage to close the gap between the theory and practice? How evolutionary algorithms do compare with operation research methods in real-world applications? Why do so few papers on evolutionary algorithms describe real-world applications? For what type of problems evolutionary algorithm is “the best” method? We’ll attempt to answer these questions – and the answers will be illustrated by examples of real-world applications of evolutionary algorithms.

Presenter Biography:
Zbigniew Michalewicz is Professor of Computer Science at the University of Adelaide in Australia. He completed his Masters degree at Technical University of Warsaw in 1974 and he received Ph.D. degree from the Institute of Computer Science, Polish Academy of Sciences, in 1981. He also holds a Doctor of Science degree in Computer Science from the Polish Academy of Science. Zbigniew Michalewicz also holds Professor positions at the Institute of Computer Science, Polish Academy of Sciences, the Polish-Japanese Institute of Information Technology, and the State Key Laboratory of Software Engineering of Wuhan University, China. He is also associated with the Structural Complexity Laboratory at Seoul National University, South Korea.  He is currently Professor of Computer Science at the University of Adelaide in Australia.
Zbigniew Michalewicz is the Chairman and Chief Scientific Officer of SolveIT Software.  SolveIT Software Pty Ltd is an Australian company specialising in advanced planning & scheduling, supply & demand optimisation, predictive modelling and mining solutions. They provide mining consulting services and mining software applications covering mining exploration management, pit to port logistics, mining simulation software  and mine planning software. Customers include Rio Tinto Iron Ore, Rio Tinto Simandou, Xstrata Coal, Xstrata Copper, Xstrata Zinc, BHP Billiton Iron Ore, BMA Coal, Fortescue Metals Group, Hancock Prospecting and Pacific National Coal.
Zbigniew Michalewicz has published over 200 articles and 15 books on subjects related to heuristic methods. These include Adaptive Business Intelligence and How to Solve It: Modern Heuristics (both published by Springer). Other books include a monograph Genetic Algorithms + Data Structures = Evolution Programs (3 editions, a few translations), Winning Credibility: A guide for building a business from rags to riches, where he described his business experiences over the last years, and the most recent: Puzzle-based Learning: An Introduction to Critical Thinking, Mathematics, and Problem Solving.

Probabilistic Graphical Approaches for Learning, Modeling, and Sampling in Evolutionary Multi-objective Optimization

KC Tan

Abstract:  Multi-objective optimization is widely found in many fields, such as logistics, economics, engineering, or whenever optimal decisions need to be made in the presence of trade-offs between two or more conflicting objectives. The incorporation of probabilistic graphical approaches in evolutionary mechanism may enhance the iterative search process when interrelationships of the archived data has been learned, modeled, and used in the reproduction for multi-objective optimization. This talk will discuss the implementation of probabilistic graphical approaches in solving multi-objective optimization problems under the evolutionary paradigm. First, the problem of multi-objective optimization and its challenges such as complexities in terms of multimodal, high-dimensional, epistatic, deceptive, constrained, and uncertainties etc., will be studied. It will then show that the optimization problem can be solved using probabilistic graphical models. A binary stochastic neural network, named restricted Boltzmann machine (RBM), will be applied, and its learning, modeling and sampling mechanisms will be highlighted. A few case studies on implementing RBM for solving multi-objective optimization problems and detailed examination of how and what information are learned and modeled by the RBM will be presented. The issues of scalability and uncertainty will also be discussed in this talk.

Presenter Biography:
Kay Chen TAN is currently an Associate Professor in the Department of Electrical and Computer Engineering, National University of Singapore. He is actively pursuing research in computational and artificial intelligence, with applications to multi-objective optimization, scheduling, automation, data mining, and games.

Dr Tan has published over 100 journal papers, over 100 papers in conference proceedings, co-authored 5 books including Multiobjective Evolutionary Algorithms and Applications (Springer-Verlag, 2005), Modern Industrial Automation Software Design (John Wiley, 2006; Chinese Edition, 2008), Evolutionary Robotics: From Algorithms to Implementations (World Scientific, 2006; Review), Neural Networks: Computational Models and Applications (Springer-Verlag, 2007), and Evolutionary Multi-objective Optimization in Uncertain Environments: Issues and Algorithms (Springer-Verlag, 2009), co-edited 4 books including Recent Advances in Simulated Evolution and Learning (World Scientific, 2004), Evolutionary Scheduling (Springer-Verlag, 2007), Multiobjective Memetic Algorithms (Springer-Verlag, 2009), and Design and Control of Intelligent Robotic Systems (Springer-Verlag, 2009).

Dr Tan is currently a Distinguished Lecturer of IEEE Computational Intelligence Society. He has been invited to be a keynote/invited speaker for over 20 international conferences. He served in the international program committee for over 100 conferences and involved in the organizing committee for over 30 international conferences, including the General Co-Chair for IEEE Congress on Evolutionary Computation 2007 in Singapore and the General Co-Chair for IEEE Symposium on Computational Intelligence in Scheduling 2009 in Tennessee, USA.

Dr Tan is currently the Editor-in-Chief of IEEE Computational Intelligence Magazine (5-Year IF: 4.094; IF: 2.833 - Rank 13 out of all 127 IEEE journals). He also serves as an Associate Editor / Editorial Board member of over 15 international journals, such as IEEE Transactions on Evolutionary Computation, IEEE Transactions on Computational Intelligence and AI in Games, Evolutionary Computation (MIT Press), European Journal of Operational Research, Journal of Scheduling, and International Journal of Systems Science.

Dr Tan is the awardee of the 2012 IEEE Computational Intelligence Society (CIS) Outstanding Early Career Award for his contributions to evolutionary computation in multi-objective optimization. He also received the Recognition Award (2008) from the International Network for Engineering Education & Research (iNEER) for his outstanding contributions to engineering education and research. He was also a winner of the NUS Outstanding Educator Awards (2004), the Engineering Educator Awards (2002, 2003, 2005), the Annual Teaching Excellence Awards (2002, 2003, 2004, 2005, 2006), and the Honour Roll Awards (2007). Dr Tan is currently a Fellow of the NUS Teaching Academic.

Fuzzy-IEEE 2012 Invited Lectures

The quest for transitivity, a showcase of fuzzy relational calculus

Bernard De Baets

Abstract: We present two relational frameworks for expressing similarities and preferences in a gradual way. The main focus is on the occurrence of various types of transitivity. The first framework is that of fuzzy relations; the corresponding transitivity notion is parametrized by a conjunctor, most often a triangular norm. We discuss two approaches to the measurement of similarity of fuzzy sets: a logical approach based on bi-residual operators and a cardinal approach based on fuzzy set cardinalities. In the latter approach a key role is played by analogues of the classical Bell inequalities. Here, we point out the link with the Poincaré paradox. The second framework is the cycle-transitivity framework for reciprocal relations; the corresponding transitivity notion is parametrized by an upper bound function. It plays a crucial role in the description of different types of transitivity arising in the comparison of random variables, such as in the case of winning probabilities among a set of dice or the mutual rank probability relation of a partially ordered set. We pay ample attention to the occurrence of cycles and point out the link with the ancient game of Rock-Paper-Scissors (RPS), an increasingly popular metaphor in science.

Presenter Biography:
Bernard De Baets holds an M.Sc. degree in Mathematics (1988), a Postgraduate degree in Knowledge Technology (1991) and a Ph.D. degree in Mathematics (1995). He is a Full Professor (2008) in Applied Mathematics at Ghent University, Belgium, where he is leading the research unit Knowledge-based Systems (KERMIT, 2000). He was a Government of Canada Award holder (1988-89) at the University of Saskatchewan (Canada). He is an affiliated professor (2009) at the Anton de Kom University (Suriname) and an Honorary Professor (2006) of Budapest Tech (Hungary). He was elected Fellow of the International Fuzzy Systems Association (IFSA) in 2011. KERMIT is an interdisciplinary team of (bio-)engineers, computer scientists and mathematicians. Its current activities consist of three interwoven threads: knowledge-based, predictive and spatio-temporal modelling. Bernard De Baets has acted as supervisor of 32 Ph.D. students. At present, numerous Ph.D. students are involved in the research activities of KERMIT, either in-house, through affiliations or in the framework of joint projects. Due to its unique position, KERMIT serves as an attraction pole for applications in the applied biological sciences. The bibliography of Bernard De Baets comprises more than 300 publications in international peer-reviewed journals, 60 chapters in books and 260 contributions to proceedings of international conferences, for which he received numerous best paper awards. He delivered nearly 200 lectures at conferences and research institutes. B. De Baets is co-editor-in-chief (2007) of Fuzzy Sets and Systems and member of the editorial board of several other journals.

Cognition-inspired fuzzy modelling

Maria Rifqi

Abstract: This talk presents different notions used for fuzzy modelling that formalize fundamental concepts used in cognitive psychology and in cognition. From a cognitive point of view, the tasks of categorization, pattern recognition or generalization lie in the notions of similarity (Tversky, 77), resemblance (Wittgenstein, 53) and prototypes (Rosch, 78). The same tasks are crucial in Artificial Intelligence to reproduce human behaviors. As most real world concepts are messy and open-textured (Rissland, 06), fuzzy logic and fuzzy set theory can be the relevant framework to model all these key notions.

On the basis of the essential works of Rosch and Tversky, we study a formal approach of the notions of similarity, typicality and prototype, thanks to fuzzy set theory. We propose a framework to understand the different properties and possible behaviors of different families of similarities. We highlight their semantic differences and we propose numerical tools to quantify these differences, considering different views: global/local view, order-based/value-based comparisons. A platform of comparisons of similarity measures will be presented.

The construction of fuzzy prototypes lead us to a new supervised learning method. Its use in several different applications like medical applications, emotion recognition from physiological signals or from texts, will be presented.

Presenter Biography:
Dr. Maria Rifqi has been an associate professor at the University Panthéon-Assas and with LIP6, the computer science laboratory of the University Pierre et Marie Curie, since 1998. She obtained a MSc degree in computer science and operational research in 1993 and a PhD in Computer Science in 1996 as well as her Habilitation in Computer Science in 2010 from the University Pierre et Marie Curie.

She was the scientific chair of LFA 2010, the french conference on fuzzy logic and its applications. She was also the special sessions chair of FUZZ'IEEE 2010 and the local arrangement chair of SSCI 2011. She participated to several international conference program committee: ECSQARU, IPMU, FlexDBIST, FQAS, EUSFLAT, IEEE/ACM CSTST. She is the machine learning and data mining area editor of the International Journal of Uncertainty and Fuzziness in Knowledge Based Systems.

She has published around 60 papers in international journals or conferences and books. Her research includes fuzzy logic, approximate reasoning, and prototypes learning. She is specialized in similarity measures and machine learning algorithms.
 

A Unified Fuzzy Model-Based Framework for
Modeling and Control of Complex Systems:
From Flying Vehicle Control to Brain-Machine Cooperative Control

Kazuo Tanaka

Abstract: This talk presents an overview of a unified fuzzy model-based framework for modeling and control of complex systems. A number of practical applications, ranging from flying vehicle control to brain-machine cooperative control, are discussed in detail. The theory and applications have been developed in our laboratory at the University of Electro-Communications (UEC), Tokyo, Japan, in collaboration with Prof. Hua O. Wang and his laboratory at Boston University, Boston, USA.
T
he first part of this talk gives a comprehensive treatment of advances on fuzzy control utilizing linear matrix inequalities (LMIs) and more recently sum of squares (SOS). A key feature of both approaches is that they provide simple, natural and effective design procedures as alternatives or supplements to other nonlinear control techniques that require special and rather involved knowledge. The LMI-based design approaches entail obtaining numerical solutions by convex optimization methods such as the interior point method. Though LMI-based approaches have enjoyed great success and popularity, there still exist a large number of design problems that either cannot be represented in terms of LMIs, or the results obtained through LMIs are sometimes conservative. A post-LMI framework is SOS-based approaches for modeling and control of nonlinear systems using polynomial fuzzy systems.
The second part of this talk discusses applications to real world nonlinear systems. The applications include not only robust control, guaranteed cost control, multi-objective control problems, etc., of nonlinear systems but also challenging control tasks from flying vehicle control to brain-machine cooperative control. Flying vehicle control is one of the more challenging control problems. This talk presents control of flying vehicles such as a micro helicopter, a powered paraglider, etc., via the LMI and SOS approaches. These applications feature unstable dynamics, nonlinearities, and coupling between the positions and attitudes. Furthermore, this talk touches upon a most recent result on wireless vision-based stabilization of an indoor micro helicopter via visual simultaneous localization and mapping. Although path-tracking control using only a small single wireless vision sensor is a quite challenging task, the results demonstrate the viability of our approach. In addition, this talk introduces our newly developed powered paraglider control system for rescue missions.
A
nother application of interest is the brain-machine cooperative control of an electrical wheelchair. Brain-machine interfaces (BMIs) have been developed extensively within the framework of electroencephalogram (EEG) signal classification. Today, there exists a large body of literature on brain-machine interface constructions and their applications. In particular, there has been a flurry of BMI research activities on motor imaginary, event related potentials (ERP) such as P300, and/or steady-state visual evoked potential (SSVEP), etc. The last part of this talk introduces our recent efforts on brain-machine cooperative control of an electrical wheelchair. The methodology features a well-balanced approach based on fuzzy blending of laser-range finder based feedback control and brain activity–based control on a recently developed BMI platform.

   Presenter Biography:
Kazuo Tanaka received Ph.D. degree, in Systems Science from Tokyo Institute of Technology, in 1990, respectively. He is currently a Professor in Department of Mechanical Engineering and Intelligent Systems at The University of Electro-Communications. He was a Visiting Scientist in Computer Science at the University of North Carolina at Chapel Hill in 1992 and 1993. He received the Best Young Researchers Award from the Japan Society for Fuzzy Theory and Systems in 1990, the Outstanding Papers Award at the 1990 Annual NAFIPS Meeting in Toronto, Canada, in 1990, the Outstanding Papers Award at the Joint Hungarian-Japanese Symposium on Fuzzy Systems and Applications in Budapest, Hungary, in 1991, the Best Young Researchers Award from the Japan Society for Mechanical Engineers in 1994, the Best Book Awards from the Japan Society for Fuzzy Theory and Systems in 1995, 1999 IFAC World Congress Best Poster Paper Prize in 1999, 2000 IEEE Transactions on Fuzzy Systems Outstanding Paper Award in 2000, the Best Paper Selection at 2005 American Control Conference in Portland, USA, in 2005.

He is currently serving on the IEEE Control Systems Society Conference Editorial Board. He is also an Associate Editor for IEEE Transactions on Fuzzy Systems, Automatica, etc. He is the author of two books and a co-author of 9 books. Recently, he co-authored (with Hua O. Wang) the book Fuzzy Control Systems Design and Analysis: A Linear Matrix Inequality Approach (Wiley-Interscience). His research interests include nonlinear control systems design and analysis, flying-robot and aerial-vehicle control, and brain-machine interface.

IEEE-IJCNN 2012 Invited Lectures

Predictive Learning, Knowledge Discovery and Philosophy of Science

Vladimir Cherkassky

Electrical & Computer Engineering
University of Minnesota
Minneapolis MN 55455
email: cherk001<AT>umn.edu
http://www.ece.umn.edu/users/cherkass/predictive_learning/
 

Abstract: Various disciplines, such as machine learning, statistics, data mining and artificial neural networks, are concerned with the estimation of data-analytic models. A closer inspection reveals that a common theme among all these methodologies is estimation of predictive models from data. In our digital age, an abundance of data and cheap computing power offers hope of knowledge discovery via application of statistical and machine learning algorithms to empirical data. This data-analytic knowledge has both similarities and differences with classical first-principle scientific knowledge. For example, any scientific theory can be viewed as inductive theory because it generalizes over a finite number of observations (or experiments). Thus, the problems of induction and knowledge discovery have been thoroughly investigated in Western philosophy of science. This philosophical analysis dates back to Kant and Hume who investigated the problem of logical induction, as well as psychological induction. Any knowledge involves a combination of hypotheses/ideas and empirical data. In the modern digital age, the balance between ideas (mental constructs) and observed data (facts) has completely shifted. Classical scientific knowledge was produced mainly by a stroke of genius (e.g., Newton, Maxwell, Einstein). In contrast, much of modern knowledge in life sciences and social sciences is derived via data-analytic modeling. This data-driven knowledge is obtained following the VC-theoretical methodological framework, also known as predictive learning. This paper presents a brief survey of the philosophical concepts related to inductive inference, and then extends these ideas to predictive data-analytic knowledge discovery. Further, we relate these classical philosophical ideas to modern statistical learning. Finally, we apply the philosophical and methodological framework of predictive learning to interpretation of data-analytic models, using application examples from financial engineering and life sciences.

Presenter Biography:

Vladimir Cherkassky is Professor of Electrical and Computer Engineering at the University of Minnesota, Twin Cities. He received MS in Operations Research from Moscow Aviation Institute in 1976 and PhD in Electrical and Computer Engineering from the University of Texas at Austin in 1985. He has worked on theory and applications of statistical learning since late 1980’s and he has co-authored the monograph Learning from Data, now in its second edition. He is also the author of the forthcoming book Introduction to Predictive Learning.
He has served on editorial boards of IEEE Transactions on Neural Networks (TNN), Neural Networks (the official journal of INNS), Natural Computing, and Neural Processing Letters. He was a Guest Editor of the IEEE TNN Special Issue on VC Learning Theory and Its Applications published in September 1999. Dr. Cherkassky was organizer and Director of NATO Advanced Study Institute (ASI) From Statistics to Neural Networks: Theory and Pattern Recognition Applications held in France in 1993. He received the IBM Faculty Partnership Award in 1996 and 1997 for his work on learning methods for data mining. He has been elected as Fellow of IEEE in 2007, for ‘contributions and leadership in statistical learning and neural networks’.  In 2008, he received the A. Richard Newton Breakthrough Research Award from Microsoft for ‘development of new methodologies for predictive learning’.

EvoSpike: Evolving Probabilistic Spiking Neural Networks and Neuro-Genetic Systems for Spatio- and Spectro-Temporal Data Modelling and Pattern Recognition

Nikola Kasabov

Knowledge Engineering and Discovery Research Institute - KEDRI, Auckland University of Technology,
and Institute of Neuroinformatics - INI, ETH and University of Zurich
email: nkasabov<AT>aut.ac.nz
www.kedri.info
 

Abstract: Spatio- and spectro-temporal data (SSTD) are the most common data in many domain areas, including bioinformatics, neuroinformatics, ecology, environment, medicine, engineering, economics, etc. Still there are no sufficient methods to model such data and to discover complex spatio-temporal patterns from it. The brain is functioning as a spatio-temporal information processing machine and brilliantly deals with spatio-temporal data, thus being a natural inspiration for the development of new methods for SSTD. This research aims at the development of new methods for modeling and pattern recognition of SSTD, called evolving probabilistic spiking neural networks (epSNN), along with their applications.
epSNN are built on the principles of evolving connectionist systems and eSNN in particular and on probabilistic neuronal models. The latter extent the popular leaky integrate-and-fire spiking model with the introduction of some biologically plausible probabilistic parameters. The epSNN are evolving structures that learn and adapt to new incoming data in a fast incremental way.
The research explores several approaches to creating epSNN for SSTD, from a single neuron, to reservoir computing and neuro-genetic systems. A single neuronal model can capture SSTD and it can also generate a precise spike time sequence in response to a SST pattern of spikes from hundreds and thousands of inputs/synapses. The research explores different types of neuronal models and dynamic synapses, including a SPAN model, Fusi’s algorithm implemented on the INI Zurich (www.ini.unizh.ch) SNN chip, and a novel stepSNN model that implements the time-to-first spike principle and probabilistic synapses.
The research explores further ensembles of neurons and neuronal structures that may be called ‘reservoirs’. Here they are recurrent SNN that are evolving and deep learning structures, capturing spatial- and temporal components in their interaction and integration. The epSNN spatio-temporal states can be identified and classified for pattern recognition tasks, which is illustrated through some preliminary experiments on gesture- and sign language recognition, moving object recognition, EEG data recognition. The epSNN can learn data in an on-line manner using a frame-based input information representation, or alternatively - an event-address based representation (EAR), the latter implemented in the INI Zurich silicon retina chip and DVS camera and the silicon cochlea chip. The project also explores how epSNN can be used to implement finite automata models and associative memories.
A main problem in the EvoSpike model and system development is the optimization of numerous parameters. For this purpose three approaches are proposed: using evolutionary computation methods; using a gene regulatory network (GRN) model, or using both in one system, depending on the application. Linking gene/protein expression to epSNN parameters may also lead to new types of neuron-synapse-astrocydes models inspired by new findings in neuroscience. Neurogenetic models are promising for modeling and prognosis of neurodegenerative diseases such as Alzheimer’s disease and for personalized medicine in general. Future research is expected to continue through tighter integration of knowledge and methods from information science, bioinformatics and neuroinformatics. The research is funded by the EU FP7 Marie Curie project, the Knowledge Engineering and Discovery Research Institute KEDRI (www.kedri.info) of the Auckland University of Technology and the Institute for Neuroinformatics, University of Zurich and ETH (INI,
www.ini.unizh.ch).

Presenter Biography:

Professor Nikola Kasabov, FIEEE, FRSNZ is the Director of the Knowledge Engineering and Discovery Research Institute (KEDRI), Auckland. He holds a Chair of Knowledge Engineering at the School of Computing and Mathematical Sciences at Auckland University of Technology. Currently he is an EU FP7 Marie Curie Visiting Professor at the Institute of Neuroinformatics, ETH and University of Zurich. Kasabov is a Past President of the International Neural Network Society (INNS) and also of the Asia Pacific Neural Network Assembly (APNNA). He is a member of several technical committees of IEEE Computational Intelligence Society and a Distinguished Lecturer of the IEEE CIS. He has served as Associate Editor of Neural Networks, IEEE TrNN, IEEE TrFS, Information Science, J. Theoretical and Computational Nanosciences, Applied Soft Computing and other journals. Kasabov holds MSc and PhD from the Technical University of Sofia, Bulgaria. His main research interests are in the areas of neural networks, intelligent information systems, soft computing, bioinformatics, neuroinformatics. He has published more than 450 publications that include 15 books, 130 journal papers, 60 book chapters, 28 patents and numerous conference papers. He has extensive academic experience at various academic and research organisations in Europe and Asia. Prof. Kasabov has received the AUT VC Individual Research Excellence Award (2010), Bayer Science Innovation Award (2007), the APNNA Excellent Service Award (2005), RSNZ Science and Technology Medal (2001), and others. He is an Invited Guest Professor at the Shanghai Jiao Tong University (2010-2012). More information of Prof. Kasabov can be found on the KEDRI web site: http://www.kedri.info. ’.

Uncovering the Neural Code of Learning Control

Jennie Si

Department of Electrical Engineering, Arizona State University
email: This e-mail address is being protected from spambots. You need JavaScript enabled to view it
Website
 

Abstract: How interacting neurons give rise to meaningful behavior is an ultimate challenge to neuroscientists. To make the problem tractable, in my lab, a rat model is used to elucidate how cortical neural activities lead to conscious, goal-directed movement and control. However we allow rats to freely move about in the experimental apparatus so to capture their natural movement and mental conditions. This talk discusses findings based on single unit, multi-channel simultaneous recordings from rat’s frontal areas while they learned to perform a decision and control task. By exploring the neural activities from the rat’s cortical regions, we developed and utilized analytical techniques to uncover the interactions between neurons at different time scales. The findings provide interesting neural substrate to rat’s learning control behavior. The work involves both experimental and computational studies. In the experiment, rats were placed in a Skinner box for a self-paced lever pressing task that they learned by trial and error. The goal of the task is to switch a sided light cue to a center location from one of five locations. The movement of the light can be controlled by the rat with the press of either a left or a right lever. Our computational modeling reveals neural adaptation as rats learned to master the task. Our results entail both high level statistical snapshots of the neural data and more detailed dynamic modeling with functional synaptic efficacies to capture before and after learning neural characteristics and their relationships to behavior. While performing the analyses, we aimed at providing mecahstic account of how brains generate meaningful behaviors under our designed experimental condition using biologically plausible computational models.

Presenter Biography:

Jennie Si received her B.S. and M.S. degrees from Tsinghua University, Beijing, China, and her Ph.D. from the University of Notre Dame. She has been on the faculty in the Department of Electrical Engineering at Arizona State University since 1991. Dr. Si's research focuses on dynamic optimization using learning and neural network approximation approaches, namely approximate dynamic programming. In 2006, she started building a neuroscience lab to study the neural mechanism of adaptation and control using multichannel single unit recordings from rat’s motor cortical areas. She received the NSF/White House Presidential Faculty Fellow Award in 1995, and Motorola Engineering Excellence Award the same year. She is a Fellow of the IEEE, and a distinguished lecturer for the IEEE Computational Intelligence Society. She is past Associate Editor of the IEEE Trans. on Semiconductor Manufacturing; IEEE Trans. on Automatic Control, and IEEE Trans. on Neural Networks. She is current Action Editor of Neural Networks. Dr. Si has served on several professional organizations' executive boards and international conference committees. She is the Vice President for Education in the IEEE Computation Intelligence Society (2009-2012). Dr. Si was an advisor to the NSF Social Behavioral and Economical directory, and served on several proposal review panels. She consulted for Intel, Arizona Public Service, and Medtronic.

 

 

 
Contents by Vinh Quang Bui, DSARC, UNSW@ADFA, Australia / Template by Next Level Design