A State-Based Sensitivity Analysis for Distinguishing the Global Importance of Predictor Variables in Artificial Neural Networks

Figure 1

Representations of artificial neural networks. (a) A multilayer perceptron depicting interaction/influences among input variables (), hidden/output layers with processing elements ( and ), synaptic weights (), and modeled output (). (b) A representative network interpretation diagram for a representative simulation within Case Study. Line thickness portrays the relative magnitude of the weight.

(a)

(b)