Professor Marie Cottrell, Université Paris 1 Panthéon-Sorbonne
Theoretical and applied aspects of the self-organizing maps
The SOM is widely used, easy to implement, has nice properties for data mining by providing both clustering and visual representation. It acts as an extension of the k-means algorithm that complies as much as possible with the topological structure of the data.
However, since its origin, the mathematical study is difficult and is completed only in very special cases. In WSOM 2005, Jean-Claude Fort presented the state of the art, the main remaining difficulties and the mathematical tools that can be mobilized: Markov chains, theory of Ordinary Differential Equations, the theory of stability, … In this lecture we try if possible to present the theoretical advances made since then.
In addition, we pass in review some of the many original SOM algorithm variants which were defined to overcome the theoretical difficulties and/or adapt the algorithm to the processing of complex data such as time series, missing values in the data, nominal data, textual data.
Biosketch
Prof. Marie Cottrell was born in Béthune (France). She was a student at the Ecole Normale Supérieure de Sèvres, and received the Agrégation de Mathématiques degree in 1964 ( with 8° place), and the Thèse d’Etat (Modélisations de réseaux de neurones par des chaînes de Markov et autres applications) in 1988.
From 1964 to 1967, she was a High School Teacher. From 1967 to 1988, she was successively an Assistant and an Assistant Professor at the University of Paris and at the University of Paris-Sud (Orsay), except from 1970 to 1973, on which she was a Professor at the University of Habana, Cuba. From 1989 to 2012, she was a full Professor at the University Paris 1 – Panthéon-Sorbonne. Since 2012, she is a Professor Emeritus.
She is a permanent invited professor at the University of Havana, a senior member of the IEEE Society, an honorary member of the Cuban Society of Mathematics and Computation, an Honorary Doctor of the Aalto University School of Science and Technology.
Her research interests include stochastic algorithms, large deviation theory, machine learning and computational intelligence, theory and algorithms for artificial neural networks, linear and non linear exploratory data analysis, non linear forecasting methods. Since 1986, her main work deals with artificial and biological neural networks, Kohonen maps and their applications in data analysis.
She is the author of about 140 publications in this field.. She is regularly solicited as referee or international conference program committee member. She organized several international conferences. She was in charge of the laboratory SAMM (Statistics, Analysis, Multidisciplinary Modelization) at the University Paris 1 until 2012.
Professor Pablo Estévez, University of Chile and Millennium Institute of Astrophysics, Chile
Big Data Era Challenges and Opportunities in Astronomy – How SOM/LVQ Related Machine Learning Can Contribute?
Astronomy is facing a paradigm shift caused by the exponential growth of the sample size, data complexity and data generation rates of new sky surveys. For example, the Large Synoptic Survey Telescope (LSST), which will begin operations in northern Chile in 2022, will generate a nearly 150 Petabyte imaging dataset. The LSST is expected not only to improve our understanding of time varying astrophysical objects, but also to reveal a plethora of yet unknown faint and fast-varying phenomena. To cope with a change of paradigm to data-driven astronomy new computational intelligence, machine learning and statistical approaches are needed. In this talk I will present big data era challenges and opportunities in astronomy from the point of view of machine learning. In particular, I will address the question of how SOM/LVQ related methods can contribute to cope with these challenges.
Biosketch
Risto Miikkulainen is a Professor of Computer Science at the University of Texas at Austin and a Fellow at Sentient Technologies, Inc. He received an M.S. in Engineering from the Helsinki University of Technology (now Aalto University), Finland, in 1986, and a Ph.D. in Computer Science from UCLA in 1990. Risto’s research focuses on biologically inspired computation such as neural networks and genetic algorithms: On one hand, the goal is to develop intelligent artificial systems that learn and adapt by interacting with the environment, and on the other, obtain insights into biological information processing. His three main focus areas are: (1) Neuroevolution, i.e. evolving neural networks with genetic algorithms for sequential decision tasks such as those in robotics, games, and artificial life; (2) Cognitive Science, i.e. models of natural language processing, memory, and learning that, in particular, shed light on disorders such as schizophrenia and aphasia; and (3) Computational Neuroscience, i.e. development, structure, and function of the visual cortex and episodic memory. Risto is an author of over 350 articles in these research areas, and the books “Computational Maps in the Visual Cortex” (Springer, 2004), “Lateral Interactions in the Cortex: Structure and Function” (co-edited electronic book, 1996), and “Subsymbolic Natural Language Processing” (MIT Press, 1993).
Several of his papers have won best paper awards; he also won a deployed application award from AAAI-2013 for a machine learning system to predict graduate admissions, and the BotPrize award in 2012 for creating a video game bot that passes a Turing Test. His laboratory has produced numerous software packages, including the Topographica simulator for cortical maps, the NERO machine learning game, and its open-source version for AI education, OpenNERO. Risto has served as a program chair for IEEE Computational Intelligence in Games conference and IEEE Congress on Evolutionary Computation, and as an action editor for IEEE Transactions on Computational Intelligence and AI in Games, IEEE Transactions on Autonomous Mental Development, the Machine Learning journal, the Neural Networks journal, and Journal of Cognitive Systems Research. He is currently serving in the Board of Directors of the International Neural Network Society, and is an IEEE CIS Distinguished Lecturer in 2015-2017 on topics related to evolution of neural networks and computational intelligence in games.
Professor Risto Miikkulainen, University of Texas at Austin
Using SOMs to gain insight into human language processing
While SOMs are commonly used as a tool for data visualization and data analysis, they can also serve as a model for cognitive functions in humans. Such functions include semantic and episodic memory, vision, and language. In this talk I will review how elements of sentence meaning can be laid out on a map, resulting in human-like graded semantic understanding instead of a single parse tree. I will also describe a model of the lexicon that can be fit to the individual patient with aphasia, and used to predict optical rehabilitation treatments.
Biosketch
Pablo A. Estévez (M’98–SM’04) received his professional title in electrical engineering (EE) from Universidad de Chile in 1981, and the M.Sc. and Dr.Eng. degrees from the University of Tokyo, Japan, in 1992 and 1995, respectively. He is a Full Professor with the Electrical Engineering Department, Universidad de Chile, and former Chairman of the EE Department in the period 2016-2010.
Prof. Estévez is one of the founders of the Millennium Institute of Astrophysics (MAS), Chile, which was created in January 2014. He is currently leading the Astroinformatics/Astrostatistics group at MAS. He has been an Invited Researcher with the NTT Communication Science Laboratory, Kyoto, Japan; the Ecole Normale Supérieure, Lyon, France, and a Visiting Professor with the University of Tokyo.
Prof. Estévez is the President of the IEEE Computational Intelligence Society (CIS) for the period 2016-2017. He has served as IEEE CIS Vicepresident of Members Activities, Member-at-Large of the IEEE CIS ADCOM, CIS Distinguished Lecturer and as an Associate Editor of the IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS.
Prof. Estévez was general chair of the Workshop on Self-organizing Maps 2012, and he is serving as conference chair of the International Joint Conference on Neural Networks (IJCNN), to be held in July 2016, in Vancouver, Canada.
His current research interests include neural networks, self-organizing maps, information theoretic-learning, time series analysis, and advanced signal and image processing. One of his main topics of research is the application of computational intelligence techniques to astronomical datasets and biomedical signals.
