Electronic nosedata analysis for detection of mai
- 上传人: 上海科学仪器有限公司 |大小:206.48KB|浏览:3443次|时间:2018-08-16
- 文档简介
- 您可能感兴趣的资料
-
Mobility measurements by phase analysis
-
Analysis of Greenhouse Gases
-
Analysis of Glycerine
-
Analysis of Urinary Organic Acids
-
Food Analysis by 300-MS
-
Residue Pesticides Analysis by Bruker 320-MS
-
Extraction optimization and analysis of monosacchari
-
Stacking and quantitative analysis of lovastatin in
-
Analysis of acyclovir by high performance capillary
-
A facile and sensitive chemiluminescence detection o
an!°electronicnose!±hasbeenusedforthedetectionofadulterationsofsesameoil.thesystem,comprising10metaloxidesemiconductsensors,wasusedtogenerateapatternofthevolatilecompoundspresentinthesamples.priortodifferentsupervisedpatternrecognitiontreatments,featureextractiontechniqueswereemployedtochooseasetofoptimaldiscriminantvariables.principalcomponentanalysispca,fisherlineartransformationflt,stepwiselineardiscriminantanalysisstep-lda,se-lectionbyfisherweightssfwwereused,respectively.andthen,lineardiscriminantanalysislda,probabilisticneuralnetworkspnn,backpropagationneuralnetworksbpnnandgeneralregressionneuralnetworkgrnnwereappliedaspatternrecognitiontechniquesfortheelectronicnose.asforldaandpnn,fltwasthemosteffectivefeatureextractionmethod,whilestep-ldawasthemosteffectivewayforbpnnandfltwasmoresuitableforgrnn.withonlyonesamplemisclassi?edinourexperiment,ldaismorepowerfulthanpnn.excellentresultswereobtainedinthepredictionofpercentageofadulterationinsesameoilbybpnnandgrnn.aftertrainingforsometime,bpnncouldpredicttheadulterationquantitativelymorepreciselythangrnn,whereaswithfltasitsfeatureextractionmethodandwithoutiterativetraining,grnncouldalsoyieldratheracceptableresults.?2006elsevierb.v.allrightsreserved.
标签: