【正文】 9 3 ． 2 ． 5 ．評(píng)估????????????????????????????? l 3 ． 3 本章小結(jié)??????????????????．．．?????????? 19 III 目錄 4 挖掘建模???????????????????????。 tax tobesolvedforthecurrent problem system． data isa to the Asa method， Data technologyanalyze hiigh． 1evelanalysis Mining of astorevealthe hiddenbehindthedata． ItCan amount large data,SO knowledge thebehaviorof fromanumberoftaxdatainthe of taxauthorities analysis help thecharacteristicof enterprises，lock information， identify escape tax． payers reduceauditcosts． also objectives， and inspection in onthedataresourcesofthetaxinformation This relies system mainly paper thecurrentinternationaladvanceddata tool Henan on miningSAS， province， based datathe tax data the of system． There techniques， toexplore provincial usingmining ofthe is thedata aretwoachievements： one enterprise，using analyzing gofled methodtoset the decisiontreeand regressionmodeling up logisticregression logistic is risk ofthe other modeltoscorethe mining managemententerprise． Theusing to suchascluster andassociationrulesanalyze technology analysis， outlieranalysis buildtherules of intheHenan industry the enterprise province， to library acquiring tax set the falsevalueaddedinvoices， which also finallyup trade， and positioning onthemore modelof out ． Attheendofthe focus possible writing paper,we falsely directionofdata inthetax mining system． tree， hierarchical rules，SAS Keywords： datamining． decision clustering， association II l 2 2 ． 3 ． 3 常用統(tǒng)計(jì)量????????????????????????? 13 2 ． 4 本章小結(jié)???????????????????????????? 15 3 SAS／ EM 數(shù)據(jù)挖掘方法論????????????????。在文章的最后， 重點(diǎn)就稅務(wù)系統(tǒng)今后可能的挖掘方向進(jìn)行了探討。 本文依托河南省局稅收信息系統(tǒng)的各項(xiàng)數(shù)據(jù)資源，基于目前國際先進(jìn)的數(shù) 據(jù)挖掘分析工具 SAG，利用數(shù)據(jù)挖掘技術(shù)，對(duì)省局稅務(wù)系統(tǒng)的數(shù)據(jù)挖掘工作進(jìn)行 了初步探索。 數(shù)據(jù)挖掘作為是一種深層次的數(shù)據(jù)分析方法，就是對(duì)大量的數(shù)據(jù)進(jìn)行分析， 從而揭示數(shù)據(jù)背后隱藏的知識(shí)的技術(shù)。（碩士學(xué)位論文）基于 SAS 系統(tǒng)的稅務(wù)數(shù)據(jù)挖掘研究 A to thesissubmitted University Zhengzhou forthe ofMaster degree TaxData BasedonSAS Researchon Mining System Man By Zheng Shi Supervisor： Prof． Lei Computertechnology Information Schoolof Engineering 2020 May 隨著稅收信息化的不斷深入，綜合征管系統(tǒng)、增值稅管理系統(tǒng)、出口退稅 系統(tǒng)等稅收業(yè)務(wù)應(yīng)用系統(tǒng)全面推行，全國積累了大量的稅收征管業(yè)務(wù)數(shù)據(jù)，這 些數(shù)據(jù)是稅務(wù)系統(tǒng)的寶貴資源。如何有效地利用這些資源，為各級(jí)領(lǐng)導(dǎo)決策提 供科學(xué)合理的依據(jù)、提高經(jīng)濟(jì)稅源的分析能力、強(qiáng)化稅收征收管理的精細(xì)化管 理水平，也成為目前稅務(wù)系統(tǒng)中亟待解決的問題。它能幫助稅務(wù)機(jī)關(guān)從大量的稅務(wù)數(shù)據(jù)中 分析納稅 人的行為信息，識(shí)別出納稅人走逃或虛開的特征，鎖定稽查目標(biāo)，縮 小稽查成本。主要成果有兩個(gè)：一是針對(duì)“走逃企業(yè)數(shù)據(jù)分析 ，應(yīng)用決策樹、 logistic 回歸建模方法，建立一般納稅人走逃風(fēng)險(xiǎn)評(píng)級(jí)回歸模型；二是對(duì)河南省 內(nèi)購方企業(yè)的發(fā)票信息進(jìn)行聚類分析、孤立點(diǎn)分析和關(guān)聯(lián)規(guī)則挖掘技術(shù)，建立 行業(yè)交易規(guī)則庫，定位接受虛開的增值稅發(fā) 票，建立虛開模型。 關(guān)鍵字：數(shù)據(jù)挖掘，決策樹，系統(tǒng)聚類，關(guān)聯(lián)規(guī)則， SAS ABSTRACT ABSTRACT full tax andthe ofthetax Withthe ofthe information implementation deepening business as collectionand prehensive management applicationsystems， such addedtax taxrebate SOon． export