基于机器学习技术的旅游方式偏好研究——以南京市民为例

《基于机器学习技术的旅游方式偏好研究——以南京市民为例》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。

1、南京师范大学硕士学位论文以南京市民为例基于机器学习技术的旅游方式偏好研究一姓名：张郴申请学位级别：硕士专业：人文地理学指导教师：张树夫2009-05-21硕士学位论文:AbstractAbstractPeople'spreferencestotourstyleareofpracticalsignificanceforthedevelopmentoftourism.Inthispaper,tourstylepreferenceisinvestigatedsystematicallybyemployingtheadvancesi

2、nbothpsychologyandmachinelearningcommunities.Indetail,twotypesofvariables,namelyteamtourandself-helptour,areconsideredinthispapertocharacterizethepotentialtourists,basedonwhichthequestionnaireisdesigned・Datawithrespecttothevariablesareextractedfromthesequestionnair

3、es,andaspecialmachinelearningalgorithmcalledC4.5-rulePANEisemployedfordataanalysis.Thisalgorithmconductsatwice-learningproceduretoobtainnotonlypowerfulpredictiveabilitytopotentialtouristbutalsoexcellentcomprehensibility.Theseadvantagesenableaccuratemodelingofthenon

4、linearmappingfromthevariablescharacterizingthepotentialtouriststotheobjectiveconcept(i.e.thetourstylepreferences)andexplicitanalysisofsuchmappingfromthecomprehensiblemodel.Thisempiricalapproachwasappliedtothedatafrom305Nanjingresidents,andinterestingresultswerereac

5、hed.Theresultsshowthat:(1)income,familylifecycleandeducationlevelinbackgroundcharacteristicsvariablesmainlyaccountsforthepreferenceofthetourstyle.(2)thepersonalvaluevariablesarealsoimportantfactorsindeterminingone'stourstylepreference.(3)machinelearningtechniquetha

6、tconsiderstherelationshipofmultiple(input)variablestotheobjectiveconceptsimultaneouslybehavesfarbetterthanseparatelyconsideringthestraight-forwardstatisticalrelationshipbetweeneachvariableandtheobjectiveconceptintourismdataanalysis.(4)basedonthepredictiverulesdrawn

7、fromourmodel,eighttypicalgroupsofpeoplewhoprefereitherteamtourorself-helptourareidentifiedfromthe305Nanjingresidents,whichprovidepossiblenewopportunitiesanddirectionsfornewproductdevelopmentandmarketexploration.Keywords.Tour-stylepreference;machinelearning;Nanjingr

8、esidents硕士学位论文：口录图录图1・1旅游者旅游方式偏好研究架6图1・2论文研究框架图7图2-1生活方式、心理地图和消费者行为的关系12图2-2VALSTM生活方式细分14图2-3机器学习任务15图3-1二次学习过程19图3-2单隐含层神经网络结构19图3・3Bagging示意图2