I am using program to determine ann performance to predict transfer capability of power system. I have total 500 samples out of which 400 are used for training & remaining 100 are used for testing. But while executing attached program MSE increases with iterations. % Solve an Input-Output Fitting problem with a Neural Network % Script generated by NFTOOL % Created Wed Feb 29 12:21:38 IST 2012 % % This script assumes these variables are defined: % % p_training1 – input data. % p_training1 – target data.
% inputs = p_training1; % targets = t_training1;
% Create a Fitting Network %new_testing_training; % p_t_80_20; for i=1:300 new_training; hiddenLayerSize = 9; net = fitnet(hiddenLayerSize); %net.layers{1}.transferFcn = 'logsig'; %net.layers{2}.transferFcn = 'tansig'; % Choose Input and Output Pre/Post-Processing Functions % For a list of all processing functions type: help nnprocess net.inputs{1}.processFcns = {'removeconstantrows','mapminmax'}; net.outputs{2}.processFcns = {'removeconstantrows','mapminmax'};
% Setup Division of Data for Training, Validation, Testing % For a list of all data division functions type: help nndivide net.divideFcn = 'dividerand'; % Divide data randomly net.divideMode = 'sample'; % Divide up every sample net.divideParam.trainRatio =80/100; net.divideParam.valRatio = 10/100; net.divideParam.testRatio = 10/100;
% For help on training function 'trainlm' type: help trainlm % For a list of all training functions type: help nntrain net.trainFcn = 'trainlm'; % Levenberg-Marquardt %net.trainFcn = 'trainbr';
% Choose a Performance Function % For a list of all performance functions type: help nnperformance net.performFcn = 'mse'; % Mean squared error
% Choose Plot Functions % For a list of all plot functions type: help nnplot net.plotFcns = {'plotperform','plottrainstate','ploterrhist', … 'plotregression', 'plotfit'}; % net.trainParam.mu = 1; % net.trainParam.mu_dec = 0.8; % net.trainParam.mu_inc=1.5; % net.inputWeights{1,1}.learnParam.lr = 0.5; % net.biases{1,1}.learnParam.lr = 0.5;
% Train the Network %[net,tr] = train(net,inputs,targets); [net,tr] = train(net,p_training1,t_training1); [net,tr] = train(net,p_training2,t_training2); [net,tr] = train(net,p_training3,t_training3); [net,tr] = train(net,p_training4,t_training4); new_training; end % Test the Network % outputs = net(inputs); % errors = gsubtract(targets,outputs); % performance = perform(net,targets,outputs) inputs1=p_testing; targets1=t_testing; outputs1=net(inputs1); errors=gsubtract(targets1,outputs1); performance=perform(net,targets1,outputs1) % outputs = net(p_testing); % errors = gsubtract(t_testing,outputs); % performance = perform(net,t_testing,outputs)
% Recalculate Training, Validation and Test Performance trainTargets = targets1 .* tr.trainMask{1}; valTargets = targets1 .* tr.valMask{1}; testTargets = targets1 .* tr.testMask{1}; trainPerformance = perform(net,trainTargets,outputs1) valPerformance = perform(net,valTargets,outputs1) testPerformance = perform(net,testTargets,outputs1)
% View the Network view(net)
% Plots % Uncomment these lines to enable various plots. figure, plotperform(tr) figure, plottrainstate(tr) figure, plotfit(net,inputs1,targets1) figure, plotregression(targets1,outputs1) figure, ploterrhist(errors) op=mapminmax('reverse',outputs1,TS); %plot(mse); xlswrite('Book11.xls',op,'ann15_40','a14'); MSE error increases in validation & testing domain. What canbe done?
Best Answer