Title :
A local interaction heuristic for adaptive networks
Author :
Sandon, Peter A. ; Uhr, Leonard M.
Author_Institution :
Dept. of Math. & Comput. Sci., Dartmouth Coll., Hanover, NH, USA
Abstract :
The standard heuristic for optimization of network parameters is gradient descent. This heuristic can lead to nonoptimal terminal parameter configurations in multilayer networks. By adding a heuristic that coordinates the development of nearby parameter values, this ´local minimum´ problem can be reduced. After motivating the use of local interactions during learning, the authors present simulation results that demonstrate improved learning under the heuristic.<>
Keywords :
artificial intelligence; learning systems; optimisation; adaptive networks; artificial intelligence; connectionist networks; heuristic learning; local interaction heuristic; multilayer networks; optimization; Artificial intelligence; Learning systems; Optimization methods;
Conference_Titel :
Neural Networks, 1988., IEEE International Conference on
Conference_Location :
San Diego, CA, USA
DOI :
10.1109/ICNN.1988.23863
