Chaotic Systems: Pushing the Frontiers of Classical and Quantum Dynamics

At the University of Lethbridge, Dr Keramat Ali is working on developing new approaches in dynamics, both classical and quantum, to broaden the efficiency and scope of artificial recurrent neural networks (RNNs).  His focus ranges from nonlinear dynamics, chaos and noise in information processing to biometrics and navigation of autonomous agents. Dr Ali is internationally recognized as an expert in nonlinear dynamics and his pioneering work in the area of classical and quantum chaos has brought him to a new dimension of study in the recurrent neural network field.  His current research endeavors include ascertaining the need for chaos in achieving meaningful goals, characterizing the role of chaos and noise in pattern recognition, developing softwares for biometrics and navigation of autonomous agents, and designing quantum neural networks.

The need to study chaotic neural networks is crucial, as these networks are unavoidable and necessary in real biological systems. There is a big challenge in physics though, because there is no persistence in dynamics. A breakthrough in the area of control and synchronization of chaos in recurrent neural networks could potentially transform the development of artificial neural networks and improve the efficiency of neural network systemization in robotics applications.

Dr Ali’s research is pushing the frontiers of classical and quantum dynamics, but without access to high performance computational resources he will be unable to continue with his research breakthroughs. Attempting to systemize recurrent neural networks and artificial intelligence requires extensive computation and memory storage. Resources like MACI are a necessity for computational problems of this magnitude and to maintain this type of leading edge research in physics.

ali@hg.uleth.ca

Selected Publications

M. Andrecut and M. K Ali. Example of Robust Chaos in a Smooth Map, Europhysics Letters, in press, 2001.

M. Andrecut and M. K Ali. Chaos in a Simple Boolean Network, International Journal of Modern Physics, B 14, in press, 2001.

M. Andrecut and M. K Ali. A Simple Neural Network Approach to Invariant Image Recognition, Modern Physics Letters B, in press, 2001.

Z. Tan and M. K Ali. Associative Memory using Synchronization in a Chaotic Neural Network, International Journal of Modern Physics, C 12, No 1, in press, 2001.

Z. Tan and M. K. Ali. Pattern Recognition with Stochastic Resonance in a Generic Neural Network, International Journal of Modern Physics, C 11, No. 8, in press,  2001.

·A. Potapov and M. K Ali. Chaotic Neural Control, Physical Review E 63, in press 2001.

Potapov  and M.K. Ali. Robust Chaos in Neural Networks, Physics  Letters A 277, 310-322, 2000.

M. Andrecut and M. K Ali. Self-organizing Neural Network Model of the Sensory-Motor Mechanism, International Journal of Modern Physics, B 14, 1815-1824, 2000.