Which of the following is correct about cost$-$based classification models in machine learning?

Question $16$ options:

Costs of misclassifications have the same meanings to different stakeholders of deployed models

Cost$-$based machine learning models are not needed in practice due to the availability of precision and recall measures

Estimating actual costs of misclassifications is not a straightforward process in all domains

Machine learning tools are not capable of incorporating costs of misclassifications into account

Count$-$based and cost$-$based model evaluations often lead to the same conclusions