TY - GEN

T1 - A neural network having fewer inner constants to be trained and Bayesian decision

AU - Ito, Yoshifusa

AU - Srinivasan, Cidambi

AU - Izumi, Hiroyuki

PY - 2007

Y1 - 2007

N2 - The number of constants in a neural network, such as connection weights and threshold, to be trained may decide directly the complexity of its learning space and, consequently, impact the learning process. It is also probable that the locations of the constants are related to the complexity. In addition, a constant to be trained at the first step of the BP learning may not add to the complexity of the learning space in comparison to those to be trained at the later steps. This paper, reflecting the above perspective, proposes a one-hidden-layer neural network with less complex learning space compared to that of ordinary one-hidden-layer neural networks. In particular, we construct a one-hidden-layer neural network having fewer constants to be trained, most of which are trained at the first step of the BP training. The network has more hidden-layer units than the required minimum for approximation but the number of constants to be trained is smaller. The goal of the network is to overcome the difficulties during statistical learning with dichotomous random teacher signals. As an example, we apply it to the approximation of a Bayesian discriminant function.

AB - The number of constants in a neural network, such as connection weights and threshold, to be trained may decide directly the complexity of its learning space and, consequently, impact the learning process. It is also probable that the locations of the constants are related to the complexity. In addition, a constant to be trained at the first step of the BP learning may not add to the complexity of the learning space in comparison to those to be trained at the later steps. This paper, reflecting the above perspective, proposes a one-hidden-layer neural network with less complex learning space compared to that of ordinary one-hidden-layer neural networks. In particular, we construct a one-hidden-layer neural network having fewer constants to be trained, most of which are trained at the first step of the BP training. The network has more hidden-layer units than the required minimum for approximation but the number of constants to be trained is smaller. The goal of the network is to overcome the difficulties during statistical learning with dichotomous random teacher signals. As an example, we apply it to the approximation of a Bayesian discriminant function.

UR - http://www.scopus.com/inward/record.url?scp=51749087272&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=51749087272&partnerID=8YFLogxK

U2 - 10.1109/IJCNN.2007.4371437

DO - 10.1109/IJCNN.2007.4371437

M3 - Conference contribution

AN - SCOPUS:51749087272

SN - 142441380X

SN - 9781424413805

T3 - IEEE International Conference on Neural Networks - Conference Proceedings

SP - 2993

EP - 2998

BT - The 2007 International Joint Conference on Neural Networks, IJCNN 2007 Conference Proceedings

Y2 - 12 August 2007 through 17 August 2007

ER -