基于FP-Growth的分類規(guī)則挖掘算法及其應(yīng)用.pdf

1、青島大學(xué)碩士學(xué)位論文基于FPGrowth的分類規(guī)則挖掘算法及其應(yīng)用姓名：遲麗寧申請學(xué)位級別：碩士專業(yè)：計算機(jī)軟件與理論指導(dǎo)教師：邵峰晶20120603㈤IlllIIIllIIIlllIIIIIlllY2088218AbstractT11ispaperissupportedbytheNationalOceanicandpublicresearchtopicof“theintegrationquickaccessandintelligent

2、analysistechniquesofBohaiSeamarineheterogeneousdata’WhenusingFPGrowthalgorithmtominelargedatabases，thereisthephenomenonthattheFPGrowthalgorithmwilltraversetoomuchnodesSotheimprovedFP—Growthalgorithmisprovidednotonlyteduc

3、ethenumberofnodesofthetree，butalsothenumberofthetraversednodes，andthecostoftimeandspaceAnyNcategoryclassificationproblemcanbetransfoiinedintoatwoclassclassificationproblemSobasedontheFPGrowthalgorithm，atwo—classclassific

4、ationalgorithmisprovidedInordertoavoidthestepthattheoriginalFP—GrowthalgorithmgeneratesconditionalpaRembases，thetreeproducedbyfrequent1itemsetsdoesnotincludetheclassattributes，butaddsthecountofrelativesupport，obtainsthec

5、onditionaltree晰thclassificationrulesT11isnotonlyreducesthesizeofthetree，butalseducesthespacecostandreducesthegenerationofinvalidclassificationrulesBasedontheimprovedFPGrowthalgorithm，takeMODISimagesasthedatasource，mineth

6、edetectionrulesofMODISimages，thenpresentsthegreentidedetectionmethodbasedonknowledge嘞【egreentideinMavJulYin2008inthewatersnearQingdaoasresearchobjects，conductthecasestudiesofgreentidedetection；theresdtshowsthattheminedru