TY  - JOUR
T1  - A Novel Method of Escape from Local Minima for Elman Neural Network
AU - , Zhiqiang Zhang AU - , Zheng Tang 
JO  - International Journal of Soft Computing
VL  - 2
IS  - 4
SP  - 549
EP  - 554
PY  - 2007
DA  - 2001/08/19
SN  - 1816-9503
DO  - ijscomp.2007.549.554
UR  - https://makhillpublications.co/view-article.php?doi=ijscomp.2007.549.554
KW  - Elman Neural Network (ENN)
KW  -modified error function
KW  -local minima problem
KW  -Boolean Series Prediction Question (BSPQ)
AB  - Eleman Neural Network (ENN) have been efficient identification tool in many areas (classification and prediction fields) since they have dynamic memories. However, one of the problems often associated with this type of network is the local minima problem which usually occurs in the process of the learning. To solve this problem and speed up the learning process, we propose a method to add a term in error function which related to the neuron saturation of the hidden layer for Elman Neural Network. The activation functions are adapted to prevent neurons in the hidden layer from stucking into saturation area. We apply the new method to the Boolean Series Prediction Questions to demonstrate its validity. Simulation results show that the proposed algorithm has a better ability to find the global minimum than back propagation ENN algorithms within reasonable time.
ER  -