Closed tristankr closed 5 years ago
The end result is virtually 100%, while the dataset has a notable number of positives. At that point it does not matter. To avoid doubt, in a future example we would not want to use accuracy.
Closing here, as it's really not an issue. Feel free to open a new ticket if anything.
Why the Talos examples of binary classification with imbalanced dataset like breast cancer uses accuracy for the hyperparameter optimization?