數(shù)據(jù)挖掘工具性能比較
單擊此處編輯母版標(biāo)題樣式,,單擊此處編輯母版文本樣式,,第二級(jí),,第三級(jí),,第四級(jí),,第五級(jí),,*,*,*,數(shù)據(jù)挖掘,工具性能比較,主要數(shù)據(jù)挖掘工具,SAS公司的 Enterprise Miner,,IBM公司的 Intelligent Miner,,SPSS公司的 Clementine,,Statsoft公司的Statistica Data Miner,,DB Miner公司的 DBMiner,,NCR公司的Teradata Warehouse Miner,,Unica公司的Affinium Model,,Insightful公司的Insightful Miner,,Data Miner 公司的RIK, EDM and DMSK,,Information Discovery 公司的Data Mining Suite,,Angoss 公司的 KnowledgeSTUDIO,,Data Mining Technologies 公司的 Nuggets,,Fujitsu公司的 GhostMiner,,Oracle公司的 Darwin,,數(shù)據(jù)挖掘工具選擇指導(dǎo)原則,公司的數(shù)據(jù)挖掘需求是短期行為還是長期使用,,公司的數(shù)據(jù)挖掘經(jīng)驗(yàn)和水平,,公司的數(shù)據(jù)狀態(tài),,公司的預(yù)算,,工具的性能,,工具評判-數(shù)據(jù)存取,,,功能和特征,數(shù)據(jù)存取,,,,,帶權(quán)得分,,,,,,軟件,,,,,IBM,SAS,,,特征,Intelligent,Enterprise,SPSS,,權(quán)值,Miner,Miner,Clementine,文本文件,30%,30,30,30,EXCEL文件,10%,5,10,5,通過數(shù)據(jù)庫的NATIVE接口取得數(shù)據(jù),30%,20,25,20,ODBC/JDBC/OLEDB,30%,20,25,25,總分,100%,75,90,80,工具評判-數(shù)據(jù)處理,,,功能和特征,,,數(shù)據(jù)處理,,,,,帶權(quán)得分,,,,,,軟件,,,,,IBM,SAS,,,特征,Intelligent,Enterprise,SPSS,,權(quán)值,Miner,Miner,Clementine,基本數(shù)學(xué)變化,20%,18,20,18,數(shù)據(jù)分段,5%,5,5,5,數(shù)據(jù)整合,10%,10,10,10,數(shù)據(jù)過濾,10%,10,10,10,數(shù)據(jù)轉(zhuǎn)換,10%,10,10,10,數(shù)據(jù)編碼,10%,10,10,10,數(shù)據(jù)隨機(jī)采樣,20%,15,20,20,SQL支持,15%,15,15,15,總分,10.00%,93,100,98,工具評判-模型算法,,,,,,,,,功能和特征,模型算法,,,,,帶權(quán)得分,,,,,,軟件,,,,,IBM,SAS,,,特征,Intelligent,Enterprise,SPSS,,權(quán)值,Miner,Miner,Clementine,聚類,20%,20,16,16,分類,20%,16,20,18,統(tǒng)計(jì),10%,8,10,10,關(guān)聯(lián)分析,15%,15,15,15,相關(guān)分析,10%,10,10,10,時(shí)間序列,5%,4,5,4,值預(yù)測,20%,18,20,18,總分,100%,91,96,91,工具評判-自動(dòng)建模,功能和特征,,,,,,,自動(dòng)建模,,,,,帶權(quán)得分,,,,,,軟件,,,,,IBM,SAS,,,特征,Intelligent,Enterprise,SPSS,,權(quán)值,Miner,Miner,Clementine,模型并行性,30%,30,30,25,模型優(yōu)化,20%,18,20,18,模型間結(jié)果共享,10%,9,10,8,參數(shù)設(shè)置靈活性,40%,35,40,35,總分,100%,92,100,86,工具評判-可視化技術(shù),功能和特征,可視化技術(shù),,,,,帶權(quán)得分,,,,,,軟件,,,,,IBM,SAS,,,SPSS,,Clementine,,特征,Intelligent,Enterprise,,,權(quán)值,Miner,Miner,,2-D 圖,15%,15,12,12,3-D 圖,10%,5,8,8,樹狀顯示,10%,10,10,10,散點(diǎn)圖,10%,8,10,8,線圖,10%,10,10,8,餅圖,15%,15,15,15,ROC 圖,10%,5,10,10,Gain Lift 圖,20%,20,20,20,總分,100%,88,95,91,工具評判-其它,功能和特征,其它,,,,,帶權(quán)得分,,,,,,軟件,,,,,IBM,SAS,,,特征權(quán)值,Intelligent,Enterprise,SPSS,,,Miner,Miner,Clementine,中文支持,30%,30,30,0,過度訓(xùn)練解決,15%,8,12,10,平臺(tái)通用性,20%,18,20,20,模型代碼輸出,15%,10,12,10,用戶友好界面,20%,12,18,16,總分,100%,78,92,56,工具評判-總分,功能,總分,,,,,,,,,,,軟件,,,,,IBM,SAS,,,,Intelligent,Enterprise,SPSS,,權(quán)值,Miner,Miner,Clementine,數(shù)據(jù)存取,10%,75,90,80,數(shù)據(jù)處理,20%,93,100,98,模型算法,30%,91,96,91,自動(dòng)建模,10%,92,100,86,可視化,15%,88,95,91,其它,15%,78,92,56,總分,100%,88,96,86,