Volume 10 - Issue 9
Construction final account prediction scheme based on normal least squares support vector regression network
Abstract
This paper is to present a novel construction final account prediction method based on normal least squares support vector regression network. Normal least squares support vector regression network has more robust than least squares support vector regression network. In the study, we will testify the feasibility of construction final account prediction based on normal least squares support vector regression network by the cases. The testing results show that the construction final account prediction error of normal least squares support vector regression network is smaller than that of support vector regression network. Therefore, we can conclude that the forecasting ability for construction final account of normal least squares support vector regression network is more excellent than that of traditional support vector regression network.
Paper Details
PaperID: 84901857333
Author's Name: Jia, H., Li, J.
Volume: Volume 10
Issues: Issue 9
Keywords: Construction final account, Least squares, Normal direction, Regression network
Year: 2014
Month: May
Pages: 3561 - 3567