Sensitivity ,Specificity and Accuracy calculation from Confusion matrix

I calculate correct accuracy classification from confusion matrix as 1-c ,Sensiticity TP/(TP+FN) as SE = cm(2,2)/sum(cm(2,:)) and Specificity TN/(TN+FP) SP=cm(1,1)/sum(cm(1,:)) to a binary classification problem .Are these calculations right ?I use 10 fold cross validation in my data and i want calculate CA,SE ,SP of the test set which incudes 2 samples each time.Sometimes in test all are positives and classified as positives and then specificity is NaN .How can i cope this ?Im interested for the mean of these in 10 folds .

Answer 1

[KALYAN ACHARJYA]

Dear Lea, Whether you work on Image or Numerical data sets, If you have numerical datasets and already built confusion matrix then you can easily able to compute the above parameters. Please follow the following link- http://www.dataschool.io/simple-guide-to-confusion-matrix-terminology/

Thanks, if you need more clarification, let me know here.

[fahmida afroz]

Thank you so much Mr. ACHARJYA for this documentation. I was so stressed out not having proper way.

Answer 2

[kalyani kumari]

i have specificity,sensitivity,accuracy and precision values.but unable to write the matlab code to generate average values for individual and cofusion matrix for 100 samples.plz help me out how to write the code