As other commenters have pointed out, accuracy is not a great metric to use here. A confusion matrix [1] is needed to fully capture the results. I'm surprised

As other commenters have pointed out, accuracy is not a great metric to use here. A confusion matrix [1] is needed to fully capture the results. I'm surprised that one was not prominently displayed in this article. They are standard practice for evaluating binary classifiers. [1] https://www.dataschool.io/simple-guide-to-confusion-matrix-terminology/ https://www.dataschool.io/simple-guide-to-confusion-matrix-t...