TY - JOUR
T1 - Recurrent neural networks training with stable bounding ellipsoid algorithm
AU - Yu, Wen
AU - de Jesus Rubio, José
PY - 2009
Y1 - 2009
N2 - Bounding ellipsoid (BE) algorithms offer an attractive alternative to traditional training algorithms for neural networks, for example, backpropagation and least squares methods. The benefits include high computational efficiency and fast convergence speed. In this paper, we propose an ellipsoid propagation algorithm to train the weights of recurrent neural networks for nonlinear systems identification. Both hidden layers and output layers can be updated. The stability of the BE algorithm is proven.
AB - Bounding ellipsoid (BE) algorithms offer an attractive alternative to traditional training algorithms for neural networks, for example, backpropagation and least squares methods. The benefits include high computational efficiency and fast convergence speed. In this paper, we propose an ellipsoid propagation algorithm to train the weights of recurrent neural networks for nonlinear systems identification. Both hidden layers and output layers can be updated. The stability of the BE algorithm is proven.
KW - Bounding ellipsoid (BE)
KW - Identification
KW - Recurrent neural networks
UR - http://www.scopus.com/inward/record.url?scp=67649322111&partnerID=8YFLogxK
U2 - 10.1109/TNN.2009.2015079
DO - 10.1109/TNN.2009.2015079
M3 - Artículo
C2 - 19447727
SN - 1045-9227
VL - 20
SP - 983
EP - 991
JO - IEEE Transactions on Neural Networks
JF - IEEE Transactions on Neural Networks
IS - 6
ER -