outputtest = multisvm(S​VMinput,SV​Mclass,SVM​testinput)​; % print out the output after SVM training

Dear community,

Below I attach two documents (.mat and .m) that I would like to use SVM to clasiffy the PD in power equipment.

I had change the svmtrain to fitcsvm(), and svmclassify to predict(). But there are still errors shown in the command window.

%%

Error using classreg.learning.internal.DisallowVectorOps/subsasgn (line 29)

You cannot assign to an object of class double using () indexing.

Error in multisvm (line 20)

models(k) = fitcsvm(TrainingSet,G1vAll);

Error in SwitchgearSVM_ANNonly1 (line 23)

outputtest = multisvm(SVMinput,SVMclass,SVMtestinput); % print out the output after SVM training

%%

Wish to get help from the community.

Thank you in advance.

Answer 1

[Anshika Chaurasia]

Hi Lee,

Try to replace models(k) with models{k} within multisvm function in SwitchgearSVM_ANNonly1.m.

I have attached the snippet of that part of code for your reference.

for k=1:numClasses
        %Vectorized statement that binarizes Group
        %where 1 is the current class and 0 is all other classes
        G1vAll=(GroupTrain==u(k));
        models{k} = fitcsvm(TrainingSet,G1vAll); % replace here models(k) with models{k}
    end

    %classify test cases
    for j=1:size(TestSet,1)
        for k=1:numClasses
            if(predict(models{k},TestSet(j,:))) % replace here models(k) with models{k}
                break;
            end
        end