Title :
Neural network applications to nonlinear time series analysis
Author :
Pethel, Shawn D. ; Bowden, Charles M.
Author_Institution :
US Army Missile Command, Redstone Arsenal, AL, USA
Abstract :
An algorithm for training multi-hidden-layer neural networks is presented. The algorithm system for fast training consists of a pseudoinverse matrix least squares procedure used incrementally to solve a nonlinear neural network, together with a preconditioning algorithm to preset the weights for optimum training. The system was applied to the training of chaotic time series from various standard models and compared to corresponding results published in the literature, for the same models using conventional training methods based on the method of steepest descents. In simulation, the training system was shown to obtain equivalent accuracy in a few minutes on a 80386 level PC, whereas the conventional backpropagation algorithm requires considerably more time on a CRAY supercomputer
Keywords :
learning (artificial intelligence); least squares approximations; neural nets; nonlinear control systems; time series; 80386 level PC; chaotic time series; equivalent accuracy; multi-hidden-layer neural networks; nonlinear time series analysis; preconditioning algorithm; pseudoinverse matrix least squares procedure; steepest descents; training; Chaos; Ear; Equations; Hardware; Least squares methods; Missiles; Neural networks; Supercomputers; Time series analysis; Weapons;
Conference_Titel :
Industrial Electronics, Control, Instrumentation, and Automation, 1992. Power Electronics and Motion Control., Proceedings of the 1992 International Conference on
Conference_Location :
San Diego, CA
Print_ISBN :
0-7803-0582-5
DOI :
10.1109/IECON.1992.254459