Title :
A hybrid learning method for multilayer neural networks
Author_Institution :
Dept. of Electr. Eng. & Appl. Phys., Oregon Grad. Inst. of Sci. & Technol., Beaverton, OR, USA
Abstract :
A Newton learning approach for training a multilayer neural network is provided based on an efficient derivation of Hessian matrix of the network. Since the Newton´s method converges almost quadratically, the convergence performance is improved. A hybrid learning method is developed in conjunction with the conventional backpropagation algorithm. Its performance is demonstrated by the classical XOR and parity problems
Keywords :
Hessian matrices; Newton method; backpropagation; convergence of numerical methods; learning (artificial intelligence); multilayer perceptrons; Hessian matrix; Newton learning approach; XOR problem; almost quadratic convergence; backpropagation algorithm; hybrid learning method; multilayer neural networks; parity problems; Acceleration; Backpropagation algorithms; Computational efficiency; Convergence; Information processing; Learning systems; Multi-layer neural network; Neural networks; Numerical stability; Physics;
Conference_Titel :
Neural Networks for Processing [1993] III. Proceedings of the 1993 IEEE-SP Workshop
Conference_Location :
Linthicum Heights, MD
Print_ISBN :
0-7803-0928-6
DOI :
10.1109/NNSP.1993.471888
