Posted: 2011-12-21
|
The pre-college education subcommittee organized a broad call for competition. Short youtube videos on fuzzy logic and its applications are called for with a grand prize. The competition is a success and the two winners have been selected by the competition organizing committee. The two videos are Fuzzy Egg and Fuzzy Logic: an Introduction. Check them out at http://www.youtube.com/watch?v=J_Q5X0nTmrA and http://www.youtube.com/watch?v=P8wY6mi1vV8., respectively. And forward them to your family and friends. They are produced for a general audience, not just researchers in the CIS. |
by Jim Bezdek, History Subcommittee Chair
Posted: 2011-11-21
|
Summary: IEEE Computational Intelligence Society (CIS) is committed to the collection and preservation of the history of computational intelligence, recognizing the achievements of scientists and professionals as well as inspiring future generations. The IEEE CIS History Subcommittee is focused on recording and posting the history of the development of the various forms of computational intelligence and a history of the development of the IEEE CIS. As the second installment, Jim Bezdek interviewed Past President of Neural Network Council in 1994, Patrick K. Simpson for the CIS history, made during the 2011 IEEE International Conference on Systems, Man and Cybernetics in Anchorage, Alaska, USA in October 2011. More history interviews will be made in the coming years. |
by Walter J Freeman and Robert Kozma
From tutorial presented at the 2011 International Joint Conference on Neural Networks, July 31 - August 5, 2011, San Jose, California, USA
Posted: 2011-10-24
|
Summary: Conventional digital computers store information encoded in strings of binary digits. We propose an alternative approach of pattern-based computing, in which information is stored in the form of spatial patterns of modified connections in very large-scale networks. Memories are retrieved by phase transitions, which enable cerebral cortices to construct spatial patterns of amplitude modulation of a narrow band oscillatory carrier wave. More precisely, information is encoded by spatial patterns of 'synaptic' weights of connections that couple nonlinear processing elements. Each category of sensory input that as subject can remember has a Hebbian nerve cell assembly. When accessed by a stimulus, the assembly guides cortex into a basin in a landscape of attractors, one for each category. Our approach to oscillating memory devices is strongly biologically motivated. It is based on observations that sensory information processing in the central nervous system is realized via collective oscillations of globally interacting neuronal populations. By using these robust biological mechanisms animals have capacities for perceiving and recognizing sensory signals that far surpass any existing man-made devices. This approach provides a novel view on neural networks. It includes as special cases other models, including deterministic cellular automata, such as Conway's Game of Life, Chua's cellular neural networks, mean field models like the Ising model, Amari's and Hopfield's neural network arrays, and Baars' global workspace model. In our tutorial we describe the foundational neurophysiological and computational issues that must be addressed in order to undertake bio-inspired modeling of brain dynamics and brain-machine interfacing. |
by Piero P. Bonissone, GE Research, USA
Complete talk available at
Posted: 2011-08-31
|
Summary: IEEE Computational Intelligence Society (CIS) is committed to the collection and preservation of the history of computational intelligence, recognizing the achievements of scientists and professionals as well as inspiring future generations. The IEEE CIS History Subcommittee is focused on recording and posting the history of the development of the various forms of computational intelligence and a history of the development of the IEEE CIS. As the first effort, Piero P. Bonissone interviewed Fuzzy Systems Pioneer, Jim C. Bezdek, for the CIS history, made during the 2011 IEEE International Conference on Fuzzy Systems, Taipei, Taiwan in July 2011. More history interviews will be made in the coming years. |
by George G. Lendaris, Portland State University, USA
Complete talk available at
From tutorial presented at the 2009 International Joint Conference on Neural Networks, June 14-19, Atlanta, Georgia, USA
Posted: 2011-05-18
|
Summary: Tutorial starts with basic control design problems, and brings focus on the Adaptive Critic Method for solving them. Provides a basic overview of Adaptive Critic methods, and focuses on a selected version of Adaptive Critics known as DHP. Describes the DHP process at high level, and provides intuitive description of the mathematics for developing the weight-update rules. Defines the Bellman Dynamic Programming formulation, and sets it up as foundation of the Adaptive Critic methods -- specifically the simultaneous training of the controller and the critic to optimize the Bellman Value Function. Key parameters of the resulting ADP process are described. A careful walkthrough of three (published) examples of applications of the DHP method: the traditional Pole-Cart Benchmark problem, steering controller for an autonomous 4-wheel terrestrial vehicle, and Hypersonic Aircraft control augmentation system. Details of the design process for each example are given, and provide examples of the various strategies for solving the governing equations, different examples of different Utility Functions and corresponding controllers designed by the DHP based on these Utility Functions are given, plots of experimental results for each, and conclusions drawn from each sample application. Three short video clips demonstrate different tests for the aircraft example. |
by Agostino Capponi, Purdue University, USA
Complete talk available at
From Tutorial presented at the 2009 IEEE Symposium Series on Computational Intelligence, March 30-April 2, 2009, Nashville, TN, USA
Posted: 2010-11-26
|
Summary: We start discussing the main components of credit risk frameworks which require modelling default probability, loss given default and their product which generates the credit spreads. We discuss how credit spreads are related to default risk. We review the main approaches to credit risk modeling including structural frameworks, intensity based approaches and models with incomplete information which combine the best features of the two previous approaches. We present statistical engineering methodologies used for calibrating the parameters of these models against actual market data. |
by Bruce Damer, Contact Consortium and DigitalSpace, USA
Complete talk available at
From Keynote Speech at the 2010 IEEE International Conference on Computational Intelligence and Games, August 18-21, 2010, Copenhagen, Denmark
Posted: 2010-09-02
|
Summary: After decades of developing and dreaming in the field of Artificial Life, fundamental breakthroughs in computational complexity and laboratory chemical simulation hold forth the promise that virtual worlds may become the proving grounds for an authentic artificial proto-biosystem. Early self-organization and complex phenomena within game spaces and virtual worlds suggest what form this profound new emergence might take. Key questions remain: will multi-user games and virtual worlds support the open ended autonomous evolution of AIs and ALife? If one day users encounter such entities within their worlds, how will this affect their experience, and the status of the spaces? Will games be closed to players and turned into Alife bio-preserves only open to ˇ§artificial life Biologistsˇ¨? What will the consequences be of the convergence of digital proto-biology and nano-fabrication? By the end of this Century will virtual entities walk (or swim) out of Cyberspace and into chemical reality? Biography: Bruce Damer is director of a virtual worlds research organization (the Contact Consortium) and the founder of DigitalSpace, a Silicon Valley R&D company. He was an organizer and pioneering designer of the first online virtual worlds in mid 1990s and author of the classic book ˇ§Avatars.ˇ¨ He has a decade of experience with NASA working with 3D space vehicle simulation and mission design. In 2007 Bruce instigated the EvoGrid project, a global effort to harness the power of grid computing to create conditions for the ˇ§artificial genesisˇ¨ of a digital proto-biosystem. Lastly, Bruce is co-leading a project with Stanford University and the Library of Congress to chronicle the emergence and evolution of virtual worlds, and he also curates a private collection of hundreds of vintage personal computers and game systems in his Digibarn Computer Museum. More about Bruceˇ¦s life and work can be found at http://www.damer.com. |








