文件名称：Applications of Support Vector Machines in Chemistry
文件大小：1.16MB
文件格式：PDF
更新时间：2012-10-27 09:57:53
Support Vector Machines
Kernel-basedtechniques(suchassupportvectormachines,Bayespoint machines,kernelprincipalcomponentanalysis,andGaussianprocesses)repre- sentamajordevelopmentinmachinelearningalgorithms.Supportvector machines(SVM)areagroupofsupervisedlearningmethodsthatcanbe appliedtoclassi?cationorregression.Inashortperiodoftime,SVMfound numerousapplicationsinchemistry,suchasindrugdesign(discriminating betweenligandsandnonligands,inhibitorsandnoninhibitors,etc.),quantita- tivestructure-activityrelationships(QSAR,whereSVMregressionisusedto predictvariousphysical,chemical,orbiologicalproperties),chemometrics (optimizationofchromatographicseparationorcompoundconcentrationpre- dictionfromspectraldataasexamples),sensors(forqualitativeandquantita- tivepredictionfromsensordata),chemicalengineering(faultdetectionand modelingofindustrialprocesses),andtextmining(automaticrecognitionof scienti?cinformation).