摘要
目前收入核算模型无法填补缺失的数据,模型存在核算效率低和核算错误数高的问题。提出基于大数据分析的存在可变对价的收入核算新模型构建方法,采用随机森林回归算法填补缺失的数据,改进了随机森林回归算法,在可变对价条件下填补缺失数据,在多层神经网络中输入完整的数据,实现数据的降噪处理,最后在Alpha计算框架的基础上建立收入核算新模型。实验结果表明,所提方法构建的模型核算效率高、核算错误数少。
At present,the revenue accounting model cannot fill in the missing data,and the model has the problems of low accounting efficiency and high number of accounting errors.This paper proposes a new model construction method for revenue accounting based on the existent variable consideration of big data analysis,which uses the random forest regression algorithm to fill in missing data to improve the random forest regression algorithm,fills in missing data under variable consideration conditions to input complete data in multi⁃layer neural network,realizes data noise reduction processing,and finally establishes a new revenue accounting model based on Alpha computing framework.Experimental results show that the proposed method has higher accounting efficiency and fewer accounting errors.
作者
吴永影
黄思源
WU Yong-ying;HUANG Si-yuan(Fujian Vocational College of Agriculture,Fuzhou Fujian 350007,China;Chaoyang District Staff University,Chaoyang Beijing 100020,China)
出处
《吉林工程技术师范学院学报》
2022年第7期94-98,共5页
Journal of Jilin Engineering Normal University
关键词
大数据分析
随机森林回归算法
多层神经网络
可变对价
收入核算
Big Data Analysis
Stochastic Forest Regression Algorithm
Multilayer Neural Networks
Variable Consideration
Revenue Accounting
