Abstract：Abstract ： Research purposes ： The traditional feature selection worlds，mostly based on expert experience or reference to historical statistics and analysis of engineering data to select，have a certain subjectivity. In order to accurately calculate the influence degree of each engineering feature on engineering estimation，the paper experimented witli the use of rough set theory to the engineering properties of the extraction problem，we propose a systematic construction method of rough set feature extraction. This method overcomes the subjectivity of the traditional engineering feature extraction method， plays the advantages of rough sets，thus improves accuracy and efficiency. Research conclusions：( 1 ) In this paper， the rough sets theory is used to calculate the influence degree of each engineering feature，and find out t!ie engineering characteristic categories influencing the actual engineering system. And it is more objective and true reflection of tiie actual project cost system. ( 2 ) Rough set analysis method only uses the information provided by the bill of quantities data，without any knowledge of the experience. The method is more objective than the traditional methods to determine the engineering characteristics， and the credibility is higher. (3) Using this method to extract the characteristics of subgrade engineering，compared with the conventional method of calculation，is closer to reality and tend to be safe and stable. (4) The research results can provide theoretical reference for the preparation of railway project cost，and also have certain reference significance for the municipal road and urban rail transit project estimation.
韩涛. 基于粗糙集的工程特征提取方法研究[J]. 铁道工程学报, 2017, 34(6): 77-81.
HAN Tao. Research on the Engineering Feature Extraction Metliod Based on Rough Set Theory. Journal of Railway Engineering Society, 2017, 34(6): 77-81.
2017, Vol. 34 
版权所有 © 2014 《铁道工程学报》编辑部