Research on Railway Engineering Cost Prediction Model Based on Chaotic Neural Networks and CS
It is always hard to draw on the experience of completed projects to predict engineering cost, and the nonlinear characteristic of the influence factors of engineering cost increases the difficulty of prediction. Less efforts and higher accuracy are the objects pursued by related researchers. In this paper, the Cost Significant theorem is applied to simplify computing and the chaotic neural network is used to improve accuracy. The prediction model is rooted from the nonlinear dynamic chaotic system theory and two techniques employed are phase space reconstruction and chaotic neural network construction. The experiment results indicate that the model is suitable for estimating short-term engineering investment and the prediction accuracy is improved.
Helen Zhang, Gang Shen and David Jin
Y. C. Chen "Research on Railway Engineering Cost Prediction Model Based on Chaotic Neural Networks and CS", Advanced Materials Research, Vols. 204-210, pp. 1291-1294, 2011