基于贝叶斯正则化BP神经网络的森林资源资产批量评估研究

将BP神经网络与批量评估方法结合应用于森林资源资产评估中,通过对森林资源资产评估价值主要影响因子进行数据处理,确定不同龄组的BP网络结构,结合TM算法,通过贝叶斯正则化修正BP神经网络的训练性能函数,建立基于BP神经网络的不同龄组森林资源资产批量评估模型。结果表明,贝叶斯正则化BP神经网络能够实现批量预估森林资源资产价值,其评估结果精度可靠,泛化能力很好,可为森林资源资产评估提供一种新的思路与方法。

Study on the mass appraisal of forest resources assets based on Bayesian regularization BP neural network

In this paper,BP neural network was combined with the mass appraisal method together,and this method was used to evaluate the forest resources assets.A new mass appraisal model of the forest resources in different age groups was built and this model was based on BP network structure.This structure was built by dealing with the data of main factors in evaluating the forest resources assets.In this model,the TM algorithm was introduced and the Bayesian method was used to estimate the regularization parameter and correct the BP neural network training performance function.The result showed that Bayesian regularization BP neural network model was able to evaluate the value of forest resources in batch,with an accurate result and good generalization ability.In conclude,this model could be applied to evaluate the forest resources assets.