TY - GEN
T1 - Robust artificial neural network for reliability analysis
AU - Oparaji, Uchenna
AU - Sheu, Rong-Jiun
AU - Patelli, Edoardo
PY - 2017/6/15
Y1 - 2017/6/15
N2 - Artificial Neural Networks (ANN) are used in place of expensive models to reduce the computational burden required for reliability analysis. Often, ANNs with selected architecture are trained with the back-propagation algorithm from few data representatives of the input/output relationship of the underlying model of interest. However, different performing ANNs might be obtained from the same training data, leading to an uncertainty in selecting the best performing ANN. On the other hand, using cross-validation to select the best performing ANN based on the highest R2 value can lead to a biassing in terms of the prediction made by the selected ANN. This is due to the fact that the use of R2 cannot determine if the prediction made by ANN is biased. Additionally, R2 does not indicate if a model is adequate, as it is possible to have a low R2 for a good model and a high R2 for a bad model. Hence we propose an approach to improve the prediction robustness of an ANN based on coupling Bayesian framework and model averaging technique into a unified framework. The model uncertainties propagated to the robust prediction is quantified in terms of confidence intervals. Two examples are used to demonstrate the applicability of the approach.
AB - Artificial Neural Networks (ANN) are used in place of expensive models to reduce the computational burden required for reliability analysis. Often, ANNs with selected architecture are trained with the back-propagation algorithm from few data representatives of the input/output relationship of the underlying model of interest. However, different performing ANNs might be obtained from the same training data, leading to an uncertainty in selecting the best performing ANN. On the other hand, using cross-validation to select the best performing ANN based on the highest R2 value can lead to a biassing in terms of the prediction made by the selected ANN. This is due to the fact that the use of R2 cannot determine if the prediction made by ANN is biased. Additionally, R2 does not indicate if a model is adequate, as it is possible to have a low R2 for a good model and a high R2 for a bad model. Hence we propose an approach to improve the prediction robustness of an ANN based on coupling Bayesian framework and model averaging technique into a unified framework. The model uncertainties propagated to the robust prediction is quantified in terms of confidence intervals. Two examples are used to demonstrate the applicability of the approach.
KW - Artificial Neural Network
KW - reliability analysis
KW - uncertainty quantification
U2 - 10.7712/120217.5400.17104
DO - 10.7712/120217.5400.17104
M3 - Conference contribution book
AN - SCOPUS:85043451049
T3 - UNCECOMP 2017 - Proceedings of the 2nd International Conference on Uncertainty Quantification in Computational Sciences and Engineering
SP - 651
EP - 662
BT - UNCECOMP 2017
A2 - Stefanou, George
A2 - Papadrakakis, M.
A2 - Papadopoulos, Vissarion
CY - [Athens]
T2 - 2nd International Conference on Uncertainty Quantification in Computational Sciences and Engineering, UNCECOMP 2017
Y2 - 15 June 2017 through 17 June 2017
ER -