This is really the tradeoff we face in the binary classification space. Do we want false positives or false negative reduced? There are more sophisticated appro

This is really the tradeoff we face in the binary classification space. Do we want false positives or false negative reduced? There are more sophisticated approaches, but if the costs in one direction are magnitudes more than in another, it becomes a clear preference.