how to evaluate SVM classifier using 10-fold cross validation method

how to calculate accuracy,specificity and sensitivity of SVM classifier using 10-fold cross validation method. plz can any one explain this with matlab code. Thank you,

Answer 1

[Sameer]

Hi @ANANDHI

To calculate accuracy, sensitivity, and specificity of an SVM classifier using 10-fold cross-validation :

1. Split your data into 10 folds.

2. For each fold:

• Train the SVM on 9 folds, test on the 1 remaining fold.
• Collect predictions and compare with true labels.

3. After all folds:

• Sum up true positives (TP), true negatives (TN), false positives (FP), and false negatives (FN).
• Calculate: Accuracy: ((TP+TN)/(TP+TN+FP+FN)) , Sensitivity: (TP/(TP+FN)), Specificity: (TN/(TN+FP))

Sample Example:

cv = cvpartition(Y, 'KFold', 10);
TP = 0; TN = 0; FP = 0; FN = 0;

for i = 1:cv.NumTestSets
    trainIdx = cv.training(i);
    testIdx = cv.test(i);
    SVMModel = fitcsvm(X(trainIdx,:), Y(trainIdx));
    Ypred = predict(SVMModel, X(testIdx,:));
    Ytest = Y(testIdx);
    
    TP = TP + sum((Ytest == 1) & (Ypred == 1));
    TN = TN + sum((Ytest == 0) & (Ypred == 0));
    FP = FP + sum((Ytest == 0) & (Ypred == 1));
    FN = FN + sum((Ytest == 1) & (Ypred == 0));
end

accuracy = (TP + TN) / (TP + TN + FP + FN);
sensitivity = TP / (TP + FN);
specificity = TN / (TN + FP);

fprintf('Accuracy: %.2f%%\n', accuracy*100);
fprintf('Sensitivity: %.2f%%\n', sensitivity*100);
fprintf('Specificity: %.2f%%\n', specificity*100);

For more information, Please go through the following MathWorks documentation links:

https://www.mathworks.com/help/stats/cvpartition.html

https://www.mathworks.com/help/stats/fitcsvm.html

https://www.mathworks.com/help/stats/linearmodel.predict.html

Hope this helps!