Using the Perturbed System to Analyze the Sensitivity of Influential Factors with Neural Networks

The ´perturbation´ method is an effective and widely used method in the factor sensitivity analysis in neural network design, on which two major problems: the sensitivity definition and the input perturbation ratio, are investigated in this study. Four models are considered in the investigation. Through comparison and analysis, results show that the definition derived from the partial derivatives is relatively more rational than others and, the optimum range of the input perturbation ratio could be [-20%, 20%] for a general case. Additionally, the effect of quality of model on the prediction accuracy is discussed, and their correlation is revealed.