基于二維表的關聯(lián)規(guī)則挖掘研究.pdf

1、中國民航大學碩士學位論文基于二維表的關聯(lián)規(guī)則挖掘研究姓名：孟維芬申請學位級別：碩士專業(yè)：計算機應用技術指導教師：劉山20080406中國民航大學碩上學位論文AbstractMiningforassociationrulesisanimportantclassofdataminingTheoverallperformanceofminingassociationrulesisdeterminedbyfindingallfrequentit

2、emsetsClassicminingalgorithms，suchasApriori，F(xiàn)PGrowtharepresentedApriorimaysufferfromnontrivialcosts：ItmayneedtorepeatedlyscanthedatabaseandcheckalargesetofcandidatesbypatternmatchingFPGrowthisamainmemory—basedalgorithmIt

3、sdisadvantageis：Whenthedatabaseislarge，itsFP—treecannotfitinmainmemoryTheexistingalgorithmsneglecttransactionforsupportingfrequentitemsetsAccordingtothelimitationoftheexistingalgorithm，thepaperputsforwardarithmeticnamedb

4、asedonprefixaltemationforcalculatingfrequentpatterns，includingstaticanddynamic，basedon2DimensiontableThealgorithmcompressedtransactiondatabaseintodatabaseavoidingrepeatedlyscanthedatabaseBecauseresultsincludeconcretetran

5、sactionscangetinterestingfrequentitemsetsthroughsettingtimebaseFormeetingmin_sup’SfrequentitemsetsCananalysistimeattributetogetsteadyandunsteadyclassesInthesameconditions，the20x20matrixhasmaindiagonalelements0andtheother

6、s1，F(xiàn)PGrowthalgorithmwillnotbeabletohandleThisarithmeticisusedtoanalyzethedatafromaviationaccidentAtlast，wediscussthedisadvantageofthealgorithmWiththedeepresearchofthisproject，thealgorithmwillhaveabrightfutureKeyWordsData