I have a binary classification problem.

The accuracy score is 52%

The precision for 0 is 53% and the precision for 1 is 49%

When using predict_proba() does this mean my model more accurately predicts when the outcome should be classified zero as opposed to one?

I'm not sure if this is telling me that I should be using the the first value (ynew[0][0]) returned from predict_proba() as opposed to the second (ynew[0][1]).

Here is the entire confusion matrix:

enter image description here

  • $\begingroup$ one can compare both values and decide on largest proba $\endgroup$ – Nikos M. Jan 10 at 18:34