A Study Of Professional Credit Evaluation Method Of Constructors Based On Factor Analysis And Lvq-nn

Construction has been a pillar industry of China's national economy. During 2008 to 2010, the damages caused by the products of construction increased. In order to protect people's lives and property, to produce high-quality building products, a standard and sound construction market management system is necessary needed.Sound construction market credit system can purify the environment for the developments, improve the quality of the construction market, so as to promote faster development of the construction industry. China has established a qualification system for architect, supervising engineers, cost engineers, architects, designers and other construction professors. Our law claims that large or medium size construction project must employ a professional constructor to be the project manager, so professional constructors get more focus on than the other construction professors.This thesis reviews the domestic and foreign scholars on the credit scoring. In the field of personal credit scoring, personal credit scoring used in consumer credit stays in a more mature stage. In China the study of credit for construction market (including the professional credit) has just started. In the research of credit scoring model, BP neural network is the most widely used, artificial intelligence methods, such as support vector machine approach is the recent research focus, and LVQ neural network is a specific method of classification.Combining the actual situation of construction market on the basis of academic study, this thesis select 17 evaluating indexes from 25 primary indexes by questionnaires, statistical test method for the measure of the credit scoring of constructors. Factor analysis was used to divide the 17 indicators into 5 categories, such as past working records, personal qualities, professional skills and the public credit information.Based on this index system, combined with LVQ neural network model, this thesis brings up a credit scoring method for constructors, and demonstrates the evaluation system and the scoring model is available in practice with a numerical case.