The correct answer is: B. classification problems require the output attribute to be categorical.
In classification problems, the output attribute is a category that the data can be classified into. For example, the output attribute for a spam filter might be “spam” or “not spam”. In estimation problems, the output attribute is a numeric value that the data can be estimated. For example, the output attribute for a house price prediction model might be the price of a house.
Option A is incorrect because the output attribute for classification problems can be either numeric or categorical. Option C is incorrect because classification problems do allow an output attribute. Option D is incorrect because classification problems are not designed to predict future outcomes.