2种森林资产批量评估模型的构建与比较  

Modeling and comparison of two mass appraisal models of resources assets

作　　者：赖晓燕[1] 颜桂梅[1] 宁正元[2] 陈平留[3] 

机构地区：[1]福建农林大学金山学院,福建福州350002 [2]福建农林大学计算机与信息学院,福建福州350002 [3]福建农林大学林学院,福建福州350002

出　　处：《福建农林大学学报（自然科学版）》2014年第1期39-43,共5页Journal of Fujian Agriculture and Forestry University:Natural Science Edition

摘　　要：以中龄林为例,分别建立了基于多元线性回归和BP神经网络的批量评估模型,选取测试样本对这2种模型的有效性进行检验.结果表明,所建立的2种模型对于中龄林评估都适用,但BP神经网络模型较多元线性回归模型的适用范围更广,预测精度更高.Two mass appraisal models were built for seml-mature forest, such as multiple linear regression model and BP neural net- works model. The test sample was selected to test the validity of these two models. Experimental results showed that these two mod- els were suitable for semi-mature forest assessment. BP neural networks modeI was more applicable and precise than multiple linear regression model.

关 键 词：森林资源资产评估 批量评估 多元线性回归模型 BP神经网络 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]