本文選題：納稅評(píng)估　切入點(diǎn)：廣義回歸神經(jīng)網(wǎng)絡(luò)　出處：《系統(tǒng)工程》2015年11期 　論文類型：期刊論文

【摘要】：針對(duì)上海市某區(qū)386家中小企業(yè)15個(gè)財(cái)務(wù)指標(biāo)數(shù)據(jù),運(yùn)用靈敏度分析方法篩選出對(duì)判定納稅情況具有顯著影響的10個(gè)評(píng)價(jià)指標(biāo),采用自組織神經(jīng)網(wǎng)絡(luò)方法把全部386個(gè)樣本分成性質(zhì)相似的訓(xùn)練樣本、檢驗(yàn)樣本和測(cè)試樣本,通過逐步減小光滑因子值確定其合理值,建立納稅評(píng)估廣義回歸神經(jīng)網(wǎng)絡(luò)(GRNN)模型。與線性回歸、判別分析、Logistic和支持向量機(jī)等模型的結(jié)果對(duì)比表明:GRNN模型的分類錯(cuò)誤率最低,檢驗(yàn)樣本和測(cè)試樣本的II類和I類分類錯(cuò)誤率分別低于5.4%和2.0%,平均分類錯(cuò)誤率低于2.5%.對(duì)另外339家企業(yè)納稅情況的判定結(jié)果表明,建立的GRNN模型具有很好的泛化能力和魯棒性。
[Abstract]:According to the 15 financial index data of 386 small and medium-sized enterprises in a certain district of Shanghai, the sensitivity analysis method is used to screen out 10 evaluation indexes that have a significant impact on the determination of tax payment. Using self-organizing neural network method, all 386 samples are divided into similar training samples, test samples and test samples, and the reasonable value is determined by gradually reducing the smooth factor value. A generalized regression neural network (GRNN) model for tax assessment was established. The comparison with linear regression, discriminant analysis logistic and support vector machine shows that the classification error rate of the two models is the lowest. The class II and class I classification error rates of test samples and test samples are lower than 5.4% and 2.0, respectively, and the average classification error rate is lower than 2.5.The results of tax assessment for another 339 enterprises show that the GRNN model has good generalization ability and robustness.
【作者單位】： 上海商學(xué)院財(cái)經(jīng)學(xué)院;上海理工大學(xué)管理學(xué)院;
【基金】：上海高校知識(shí)服務(wù)平臺(tái)“上海商貿(mào)服務(wù)業(yè)知識(shí)服務(wù)中心”建設(shè)子項(xiàng)目“稅收風(fēng)險(xiǎn)管理信息系統(tǒng)設(shè)計(jì)及開發(fā)”(ZF1226) 上海高校重點(diǎn)學(xué)科“商務(wù)經(jīng)濟(jì)學(xué)”建設(shè)項(xiàng)目,參加本課題的還有尹淑平、張嬌芳、史昱民、包時(shí)軍、鄔春學(xué)和高麗萍等同志
【分類號(hào)】：F812.42
，

本文編號(hào)：1592162