基于案例推理的魚病診斷專家系統(tǒng)研究.pdf

1、分類號密級‘專目差遺犬等。學(xué)位論文X774176單位代碼：10019學(xué)號：S02675基于案例推理的魚病診斷專家系統(tǒng)研究UsingCase—basedReasoningToEstablishAnExpertSystemForDiagnosisOfFishDisease研究生：周垂指導(dǎo)教師：堡疊里墼筮合作指導(dǎo)教師：申請學(xué)位門類級別：釜壟堂亟專業(yè)名稱：釜墨盤堂蘭三堡研究方向：焦!墾笠堡魚焦!墾璽綾所在學(xué)院：絲盜篷堡莖瞳二零零四年九月中圈農(nóng)業(yè)

2、人學(xué)塒!I學(xué)位論文英丘摘要AbstractInthispaperCaseBasedReasoningisputforwardtostudyfishdiseasediagnosisinordertosolvetheproblemofknowledgeacquisitionbottleneckexistinginthepreviousfishdiseasediagnosisexpertsystemsTheadvantageofCBRisfo

3、llowing：I)itjustneedsadequatecasesandtheregularizationandpretreatmentofexpertknowledgeisnotnecessary2)CBRisanefficaciousmethodtosimulateexpertthinkingwithimageryThepaperdiscussesthepresentationofanewcase，theretrievalofth

4、emostsimilarcasesfromthecasebase，andthestorageofallcasesAboveall，weanalyzethediagnosisprocessoffishdiseasethefishexperthave，andputforwardaconceptionmodeloffishdiseasediagnosisaccordingtotherelationofcaseelementsFurthermo

5、re，wedesignthecaserepresentationalformalismandthestructureofcasebaseaccordingtotherelationdatabasetheoryFishcasesarequalitativeanddiscretedata，inviewofthesecharacteristics，binarysystemisusedtepresenttheCaseswhenwediscuss

6、thecaseretrievalstrategyOnthebasisoftheabove，twocaseretrievalmethodsareprovidedoneissimilaritytestingmodel，andtheotherisroughanalysisThesimilarity—testingmodelrealizesamechanismofval‘iantweightsofthecasesaccordingtoTvers

7、ky’Scontrastmatchingmodel，Inthesecondmethod，roughanalysisisintroduced，whichcanfullyusethereductionattributesofcasedatabaseandmakenlultiindexofcaseThetwoCaseretrievaImethodsresolvetheuncertaintyoffishdiseasefromdifferentp

8、ointsofviewFinallyaccordingtothedemandofexpertsystem，themodelofJSPaddedJavaBeanshasbeendevelopedforimplementingthefishdiseasediagnosisexpertsystemunderWindows2003ServerAthreelayersstructureB／S／SusedinWEBisadoptedintilene