基于GBRT和XGBoost模型的体育场馆管理费拦标价评估  

Bidding Valuation of Stadium Management Fee Based on GBRT and XGBoost Models

作　　者：潘安娥[1] 蔡群 王瑶瑶 郑舒晴 袁珂 李怡民 尹秀娥 Pan Ane;Cai Qun;Wang Yaoyao;Zheng Shuqing;Yuan Ke;Li Yimin;Yin Xiue

机构地区：[1]武汉理工大学管理学院,武汉430000 [2]湖北中天诚资产评估有限公司,武汉430000

出　　处：《中国资产评估》2024年第6期69-80,F0003,共13页Appraisal Journal of China

摘　　要：拦标价评估是招标过程中出现的一项新型评估业务。本文以LDT体育场馆管理费拦标价评估为例,对传统的市场法在可比案例和可比因素选择及其差异修正过程中存在的问题进行改进。主要包括:根据体育场馆主楼和副楼的特点分别选取可比案例和可比因素,利用Python软件爬取大样本数据,通过灰色关联度模型分别筛选出关键特征变量;利用机器学习算法筛选高精度的估值模型,最终选用GBRT模型和XGBoost模型对拦标价进行预测,得出的结果具有较强的客观性和准确性。Bidding valuation is a new type of valuation business in the process of bidding.This paper takes LDT stadium management fee assessment as an example to improve the problems existing in the selection of comparable cases and factors and the correction of differences in the traditional marketing method.It mainly includes:selecting comparable cases and factors respectively according to the characteristics of the main building and the auxiliary building of the stadium,using Python software to obtain large sample data,and screening key feature variables respectively through the grey correlation degree model;Machine learning algorithm is used to select high-precision valuation models,and finally GBRT model and XGBoost model are selected to predict the bid price.The results obtained have strong objectivity and accuracy.

关 键 词：拦标价评估 体育场馆管理费 灰色关联度模型 机器学习 

分 类 号：F293.3[经济管理—国民经济] TP[自动化与计算机技术]