Title: Sudden risk predication model of construction supply chain based on data mining

Authors: Peng Lu; Xiaomei Li; Jinrong Nie

Addresses: School of Economic and Trade Management, Tianjin Sino-German University of Applied Sciences, Tianjin 300350, China ' School of Economic and Trade Management, Tianjin Sino-German University of Applied Sciences, Tianjin 300350, China ' Zhengzhou Shengda University of Economics, Business & Management, Zhengzhou 451191, China

Abstract: In order to improve the accuracy of quantitative evaluation of construction supply chain burst risk, improve the ability of risk prediction, and effectively guide the prevention of construction supply chain burst risk, a quantitative evaluation and prediction model of construction supply chain burst risk based on data mining and multiple regression analysis is proposed. In this method, firstly, information acquisition and adaptive feature extraction are performed to characteristic quantity in quantitative analysis of sudden risks of the construction supply chain. Secondly, a stochastic probability density model is adopted to decompose characteristics of sudden risks of the construction supply chain, and risk evaluation and relevant predication are performed to the construction supply chain through internal control and extract control. The simulation results show that the method has high accuracy in the construction supply chain sudden risk prediction, with an average prediction accuracy of 88.86%, and the shortest time cost. It can be completed in only seven seconds in the prediction process. It has good global convergence and optimisation ability in the prediction.

Keywords: data mining; the construction supply chain; risk predication; feature extraction; regression analysis.

DOI: 10.1504/IJISE.2021.118253

International Journal of Industrial and Systems Engineering, 2021 Vol.39 No.2, pp.205 - 219

Received: 24 Jun 2019
Accepted: 01 Nov 2019
Published online: 18 Oct 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article