An investigationwascarriedouton12,8

An investigationwascarriedouton12,854fortnightlytestdaymilkyieldsrecordsof
first lactationpertainingto643Sahiwalcowssiredby51bullsspreadover49years
locatedattheNationalDairyResearchInstitute,Karnal.Thecomparisonwasmade
betweentherelativeefficiencyofmultiplelinearregressionanalysisandartificial
neuralnetwork(ANN)forpredictionoffirstlactation305dmilkyield(FL305DMY)in
Sahiwalcows.ArtificialNeuralNetworkwastrainedusingthreebackpropagation
algorithmsviz. Bayesian regularization (BR), Scaledconjugategradient (SCG)and Levenberg–
Marquardt (LM). Further,thesethreealgorithmswerecomparedusingfoursetsof
trainingandtestdatasetsat66.67–33.33%,75–25%,80–20%and90–10%.Ithasbeen
foundthatthecoefficientofdeterminationofthemodelswasincreasedwiththe
additionoftestdaymilkyieldsasinputvariables.Itwasinferredfromthestudythat
artificialneuralnetworkwasbetterthanthemultiplelinearregressionanalysisto
predictFL305DMYwithmorethan80%accuracybyalmostallthemodelsatanearly
stagei.e.by111thdayofthelactationhavinglesservalueofRMSEthanMLR.Therefore,
it isrecommendedthatANNcanbeapotentialtoolforthepredictionofthefirst
lactation305-daymilkyieldinSahiwalcows.

ผลลัพธ์ (ไทย) 1:

มี investigationwascarriedouton12, 854fortnightlytestdaymilkyieldsrecordsoflactationpertainingto643Sahiwalcowssiredby51bullsspreadover49years แรกlocatedattheNationalDairyResearchInstitute,Karnal.Thecomparisonwasmadebetweentherelativeefficiencyofmultiplelinearregressionanalysisandartificialneuralnetwork (แอน) forpredictionoffirstlactation305dmilkyield (FL305DMY) ในSahiwalcows.ArtificialNeuralNetworkwastrainedusingthreebackpropagationalgorithmsviz ทฤษฎี regularization (BR), Scaledconjugategradient (เอสซีจี) และ Levenberg-Marquardt (LM) เพิ่มเติม thesethreealgorithmswerecomparedusingfoursetsoftrainingandtestdatasetsat66.67–33.33%,75–25%,80–20%and90–10% Ithasbeenfoundthatthecoefficientofdeterminationofthemodelswasincreasedwiththeadditionoftestdaymilkyieldsasinputvariables ItwasinferredfromthestudythatartificialneuralnetworkwasbetterthanthemultiplelinearregressionanalysistopredictFL305DMYwithmorethan80% accuracybyalmostallthemodelsatanearlystagei.e.by111thdayofthelactationhavinglesservalueofRMSEthanMLR.Thereforeมัน isrecommendedthatANNcanbeapotentialtoolforthepredictionofthefirstlactation305-daymilkyieldinSahiwalcows

ผลลัพธ์ (ไทย) 2:

investigationwascarriedouton12,854fortnightlytestdaymilkyieldsrecordsof
แรก กูเบส์ (BR), Scaledconjugategradient (SCG) และ Levenberg- Marquardt (LM)

ผลลัพธ์ (ไทย) 3:

การ investigationwascarriedouton12854fortnightlytestdaymilkyieldsrecordsoflactationpertainingto643sahiwalcowssiredby51bullsspreadover49years ครั้งแรกlocatedatthenationaldairyresearchinstitute karnal.thecomparisonwasmade ,betweentherelativeefficiencyofmultiplelinearregressionanalysisandartificialนิวรอลเน็ตเวิร์ค ( แอน ) forpredictionoffirstlactation305dmilkyield ( fl305dmy )sahiwalcows.artificialneuralnetworkwastrainedusingthreebackpropagationalgorithmsviz . คชกรรม regularization ( br ) , scaledconjugategradient ( SCG ) และ levenberg จำกัดมาร์คว ( LM ) เพิ่มเติม thesethreealgorithmswerecomparedusingfoursetsoftrainingandtestdatasetsat66.67 – 33.33% 75 – 25 % 80 – 20% and90 ithasbeen – 10%foundthatthecoefficientofdeterminationofthemodelswasincreasedwiththeitwasinferredfromthestudythat additionoftestdaymilkyieldsasinputvariables .artificialneuralnetworkwasbetterthanthemultiplelinearregressionanalysistopredictfl305dmywithmorethan80 % accuracybyalmostallthemodelsatanearlystagei.e.by111thdayofthelactationhavinglesservalueofrmsethanmlr.therefore ,มัน isrecommendedthatanncanbeapotentialtoolforthepredictionofthefirstlactation305 daymilkyieldinsahiwalcows .