Title :
Sensitivity analysis for feedforward artificial neural networks with differentiable activation functions
Author_Institution :
Honeywell SSDC, Minneapolis, MN, USA
Abstract :
A method for computing the network output sensitivities with respect to variations in the inputs for multilayer feedforward artificial neural networks with differentiable activation functions is presented. It is applied to obtain expressions for the first- and second-order sensitivities. An example is introduced along with a discussion to illustrate how the sensitivities are calculated and to show how they compare to the actual derivatives of the function being modeled by the neural network
Keywords :
artificial intelligence; feedforward neural nets; sensitivity analysis; differentiable activation functions; feedforward artificial neural networks; sensitivity analysis; Artificial neural networks; Backpropagation; Computer networks; Electrical equipment industry; Input variables; Multi-layer neural network; Nonhomogeneous media; Power system modeling; Process control; Sensitivity analysis;
Conference_Titel :
Neural Networks, 1992. IJCNN., International Joint Conference on
Conference_Location :
Baltimore, MD
Print_ISBN :
0-7803-0559-0
DOI :
10.1109/IJCNN.1992.287175
