Title :
Trajectory assignment by output zeroing
Author_Institution :
Lab. d´´Electron d´´Autom. et d´´Inf., Ecole des Mines d´´Ales, France
Abstract :
In this paper we investigate the problem of how to impose a state-space trajectory on an artificial neural network of Hopfield type. By defining certain output functions of network states we are able to achieve the required trajectories without requiring training. Our theory is based on nonlinear geometrical methods and we use reliable and accurate numerical algorithms for all our computations.
Keywords :
Hopfield neural nets; computational geometry; state-space methods; topology; Hopfield neural network; nonlinear geometrical methods; output zeroing; state-space trajectory; trajectory assignment; Artificial neural networks; Computer networks; Differential equations; Electronic mail; Modems; Reliability theory;
Conference_Titel :
Neural Networks, 1993. IJCNN '93-Nagoya. Proceedings of 1993 International Joint Conference on
Print_ISBN :
0-7803-1421-2
DOI :
10.1109/IJCNN.1993.713945
