Hello,
I am looking at the dice/sensitivity/specificity columns for each of the 3 labels ET, TC, and WT in the scores returned by the validation workflow. Isn't dice is supposed to be the harmonic mean of sensitivity and specificity (for each label)? But I cannot get the numbers to match up in most cases.
What am I missing?
Thanks
Darshat
Created by Darshat Shah dshah Hi Ujjwal,
I computed these as harmonic mean of sensitivity and specificity which were provided in the all_scores.csv for each class Hi Darshat,
To calculate these scores you need ground truth and the predictions by your algorithm. Considering you don't have access to the ground truths of the validation dataset, how did you compare the numbers which were mismatched?