一种基于BP神经网络的个体定税模型

An Individual Constant Tax Computing Model Based-on BP Neural Network

阐述了BP神经网络的基本原理,通过分析影响定税的一些主要因素,提出了基于BP神经网络的个体定税预测模型,并以株洲市地方税务局的系数定税数据为样本,对该模型进行分析评估,进而表明BP神经网络在核定纳税户的应纳税额方面具有较好的实用性.

The theory of back propagation neural network is explained, and a model of individual constant tax based-on BP neural network is introduced through analyzing many primary factors which effect the quantity of tax, and then the model is evaluated according to the swatch obtained from on-line database of Zhuzhou local taxation bureau. In this way, the practicability of BP neural network in constant tax is proved.