基于自适应网络模糊推理系统的旅游需求预测分析分析——以2000-2008年日本至中国月度旅游需求为例-analysis and prediction of tourism demand based on adaptive network fuzzy inference system - a case study of monthly tourism demand from japan to china from 2000 to 2008.docxVIP

的敏感性分析更贴近于实际经济社会背景。能够得到较贴合现实生活的敏感性曲线和敏感性曲面。最后，总结了文章的不足，并对今后的深入研究提出设想和建议。关键词：旅游需求；预测分析；自适应网络模糊推理系统；日本；旅游外汇收入；旅游者数量AbstractWiththerapiddevelopmentofsocietyandworldeconomy,moreandmorepeoplecomeintopursuehigh-qualitypatternoflivingaftermeetingtheirbasicstandardlifestyle.High-speeddevelopmentsofeverycountryinevitablybringabouttheboomingofworldtourismindustry.Asoneofthesomuchimportantdestinationcountries,thewholechinagainsagreatquantityofbenefits,notonlyeconomyinterestbutalsotheinterestsofthepublicandsociety.Thenwecomeintorecognizingthatthestudyonforecastinganalysisoftourismindustryissoimportanttoguaranteeefficientsustainabletourismprospect.Avarietyofforecastinganalysistechniqueshavebeendevelopedforthedeeperstudyofpredictionbyscholars.Unfortunately,allthetechniquesweusedbeforearemainlylinearequations.Wecouldnotapproachthereallyworldbyusingtraditionaltechniquesontothenon-linearrelationsoftheworld.ANN(ArtificialNeuralNetworks)isawidelyreceivedtechnique,butthepoorabilityofaccuratelyforecastingandmanydrawbacksonanalyzingtheinfluencethatfactors’impactsondependentvariableslimititsapplication.ThreeobjectiveswewanttoachieveherebyintroducingAdaptiveNetworkFuzzyInferenceSystem(ANFIS)intoforecastinganalysisontourismindustryare:Firstly,toestablishahighaccuracymodeltoprovidebasisfordecisionmakingontourismplanningandprojectmanagement;Secondly,toreducethedependenceofemployedmodelondatatocutdownthecosts.So,themodelwedevelopedherecanbeenemployedbydifferenttourismsectors;Thirdly,toenrichtheexistingforecastinganalysistechniquessothatwecanbroadentheselectivityoftourismresearchers.Bycomparingandsummarizingtheforecastinganalysistechniquescommonlyusedbefore,thepaperfindsthenecessarystudyonANFIS.Inthecasestudychapter,thepaperselectsJapanesemonthlyDemandforChina(2000.01-2008.12,108monthsintotal)asdependentvariablewhichcharacterizedbyNumberofInboundJapaneseTourist(NIT)andExpendituresofInboundTourist(EIT).Variablesaremonth(Mon),CPIsofJapan/China(CPIJ/CPIC),rateofCPIJtoCPIC(SPR),paritypriceofRMBtoJP