支持向量機在煤礦地下水位預測中的應用.pdf

1、西安科技大學碩士學位論文支持向量機在煤礦地下水位預測中的應用姓名：薛宇玲申請學位級別：碩士專業(yè)：計算機應用技術指導教師：龍熙華20060418Subject：ApplicationofSupportVectorMachinesinForecastofCoalMineUndergroundWaterTableSpecialty：ComputerApplicationTechnologyName：XueYulingInstructor：Lo

2、ngXihuaAbstract(Signature)絲Ij鱧V(Signature)魚燁Duetoheavycoalexploitationinrecentyears，manyseriousaccidentsaboutundergroundwaterofcoalminesfrequentlyhappened，whichjeopardizedtopeople’slivesandwealthSotheresearchinforecastof

3、coalmineundergroundwatertablehasactualsignificanceAforecastofcoalmineundergroundwatertablehasmanycharacteristics，suchasvariousinfluent—factorshigherdimensionnonlineartimeseriesandsoon111ispaperappliesSupportVectorMachine

4、(svM)methodtotheforecastofcoalmineundergroundwatertableSVMisanewmachinelearningtechniquedevelopedfromthemiddleof1990sbyVladimirVapnikSupportvectormachinesareaveryspecificclassofalgorithms，characterizedbytheuseofamaximalm

5、arginhyperplanethetheoryofkernels，theabsenceoflocalminima，convexoptimizationthesparsenessofthesolution，Mercer’stheoremandthecapacitycontrolobtainedbyactingonthemarginAlargenumberofexperimentshaveshownthatsupportvectormac

6、hinehasnotonlysimplerstructure，butalsobetterperformance，especiallyitsbettergeneralizationabilityFirstlythepaperstatesthetheorybasisofSVM：StatisticalLearningTheory(STL)ItmainlyintroducesthreecoreconceptsofSTL，whichareVCdi

7、mension，minimizingtheboundbyminimizinghandstructuralriskminimizationItalsoelaboratestheideas，countingstepsandoptimizealgorithmofsupportvectorclassificationandregressionSecondlyafteritanalyzesandcomparesseveraltechniqueso

8、fSVMitascertainsthesuitabletechniquesrequiredintheforecastofcoalmineundergroundwatertable，andexplainsthenecessarytheorytothemFinallythispaperestablishesamodelduringthecourseoftheforecastbyusingSVM，andanalyzestheinfluence