摘要
传统测算方法难以精准定价偏远5A级景区土地生态资源价值,如何评估其生态资源资产是推动偏远国家公园可持续发展亟需科学解决的问题。本文利用河南省焦作市修武县308个土地招牌挂样点训练机器学习测算模型,设置学习率为0.001,经过1721轮训练,均方差稳定在35.91。利用机器学习和现行方法分别对29#地价值测算,结果表明:机器学习与成交价之间方差为1.96,现行方法与成交价之间方差为14.44,方差提高12.48,机器学习测算结果与成交价仅相差1.4元/平方米,机器学习测算结果更加理想。相较传统生态资源资产定价方法,机器学习评估方法的研究对我国公园生态资源资产定价体系建设具有一定的现实意义。
Traditional measuring techniques have a tough time providing an accurate assessment of the worth of biological resources on land in distant 5A beautiful zones.The value of ecological resources is calculated using the conventional approach in this article,which is used in distant 5A picturesque places.Xiuwu County,Jiaozuo City,Henan Province,China provided the 308 land signboard samples used in this study.In the present study,a machine learning model was trained on 308 land signage sample sites in Xiuwu County,Jiazuo City,Henan Province,with a learning rate of 0.001 and an average variance of 35.91 after 1721 training rounds.According to the findings,there is a 1.96difference between transaction pricing and machine learning.The difference between the transaction price and machine learning was 1.96,while the difference between the transaction price and the present technique was 14.44,with a variance increase of 12.48.The machine learning findings are more satisfying since the price difference between them and the transaction price is only RMB 1.4 per square meter.The research of machine learning valuation techniques has some practical value for the creation of a pricing system for ecological resource assets in China’s parks as compared to the conventional way of valuing ecological resource assets.
出处
《价格理论与实践》
北大核心
2023年第1期131-134,203,共5页
Price:Theory & Practice
基金
河南省科技攻关项目(182102310950)
关键词
生态资产
地价评估
生态资源补偿
机器学习
ecological assets
land value assessment
ecological resource compensation
machine learning
