摘要
SupportVectorMachine(SVM)isapowerfulmethodologyforsolvingproblemsinnon-linearclassification,functionestimationanddensityestimation,whichhasalsoledtomanyotherrecentdevelopmentsinkernelbasedmethodsingeneral.Thispaperpresentsahighaccuracyandfault-tolerantSVMforthemobilegeo-locationproblem,whichisanimportantcomponentofpervasivecomputing.Simulationresultsshowitsbasiclocationperformance,andillustrateimpactsofthenumberoftrainingsamplesandtrainingareaontestlocationerror.
出版日期
2005年04月14日(中国期刊网平台首次上网日期,不代表论文的发表时间)