Check certain data of confusion matrix

Asked Aug 23 '17 at 12:07

Active Aug 23 '17 at 12:07

Viewed 36 times

I have a confusion matrix:

I got it by running the scikit package for confusion matrices. I would now like to find out for example the 4 False Positives and the 2 True negatives and which concrete feature values they had to make a describtive analysis. How could I do this using python? The confusion matrix is produced buy a neural net and validated. I would like somehow to find out which samples of the validation set are FP and TN to check them manually for any differences.

Thanks!

asked Aug 23 '17 at 12:07

inneb

1,060
1
9
20

please share code / sample code link that you are running. – Vikash Singh Aug 23 '17 at 12:09
2

Ideally you can do something like this `train[predictions != labels]` This will give you both `false positives` and `true negatives` – Vikash Singh Aug 23 '17 at 12:10

Check certain data of confusion matrix

0 Answers0